Analysis 1: Vitamin D analogues versus placebo

This comparison included eight vitamin D analogues for body psoriasis (see Analysis 1.1 and Table 5). Twenty trials with 3771 participants reported IAGI data on 7 of these treatments. Thirteen trials were between‐patient design, and 7 were within‐patient studies. Treatment duration ranged from 4 weeks to 12 weeks. The pooled SMD across all treatments was ‐0.95 (95% CI ‐1.17 to ‐0.74; I² statistic = 89.0%), but there was considerable variation between treatments, so we removed pooling across subgroups. Six treatments were significantly more effective than placebo, with the effect size ranging from ‐0.67 (becocalcidiol twice daily) to ‐1.66 (paricalcitol once daily). There was considerable between‐study variation in the IAGI SMD for calcitriol. The pooled effect was ‐1.03 (95% CI ‐1.71 to ‐0.36), but this ranged from ‐0.26 (95% CI ‐0.99 to 0.47) for Langner 2001 (P) to ‐3.11 (95% CI ‐3.57 to ‐2.66) for Perez 1996. The magnitude of the IAGI SMD for the Perez study was the highest across all comparisons and treatments. For the 'combined end point' of this analysis, we explored the impact of removing this trial from the pooled findings using sensitivity analysis. The presence of considerable heterogeneity within this subgroup means that the estimated average benefit should be interpreted with caution (Higgins 2011).

Analysis 2: Corticosteroid (potent) versus placebo

This comparison included 10 potent corticosteroids for body psoriasis (see Analysis 2.1 and Table 6), although no effectiveness data were available for budesonide. Nine studies with 1867 participants reported IAGI data on 6 of these 10 treatments. Eight trials were between‐patient design, and one was a within‐patient study (Stein 2001). Treatment duration ranged from 3 to 12 weeks. The SMD across all 6 treatments for IAGI was ‐1.00 (95% CI ‐1.18 to ‐0.82; I² statistic = 57.6%). All six treatments performed statistically significantly better than placebo.

Analysis 3: Corticosteroid (very potent) versus placebo

This comparison included three very potent corticosteroids for treatment of psoriasis of the body (see Analysis 3.1 and Table 7). Five studies with 515 participants reported IAGI data on 2 of the 3 treatments. There were four between‐patient trials and 1 within‐patient study (Beutner 2006). Treatment duration ranged from two to four weeks. The IAGI SMD across both treatments was ‐1.87 (95% CI ‐2.38 to ‐1.36; I² statistic = 78.7%). Both clobetasol propionate and halobetasol performed statistically significantly better than placebo. In Analysis 18.1, we present placebo‐controlled scalp trials of very potent corticosteroids.

Analysis 4: Dithranol versus placebo

Our review did not identify any study comparing dithranol against placebo and which reported IAGI data.

Analysis 5: Vitamin D combination products versus placebo

This comparison included treatment with combined calcipotriol and betamethasone dipropionate used either once or twice daily on the body (see Analysis 5.1 and Table 8). Five parallel‐group studies with 2058 participants contributed data on both dosing options. Treatment duration ranged from four to eight weeks. The IAGI SMD across treatments was ‐1.44 (95% CI ‐1.76 to ‐1.12; I² statistic = 89.4%), with twice‐daily combination treatment (SMD ‐1.90; 95% CI ‐2.09 to ‐1.71) achieving a significantly larger effect than once‐daily treatment (SMD ‐1.21; 95% CI ‐1.50 to ‐0.91). However, the difference between once‐ and twice‐daily dosing was not statistically significant when benefit was assessed using the PASI (see Analysis 5.3). In Analysis 18.1, we report placebo‐controlled trials of combination vitamin D/steroid treatments for scalp psoriasis.

Analysis 6: Other treatment versus placebo

This comparison comprised all other treatments for psoriasis of the body not included in the first five analyses; therefore, we removed pooling. None of the studies assessed the same treatment, which means that findings should be interpreted with caution.

In total, we included 26 treatments in this analysis (see Analysis 6.1 and Table 9). Eight studies with 364 participants reported IAGI data on 8 of these 26 treatments. Four trials were between‐patient design, and four were within‐patient studies. Treatment duration ranged from 3 to 12 weeks.

Four treatments performed statistically significantly better than placebo: anti‐IL‐8 monoclonal antibody cream; betamethasone 17‐valerate 21‐acetate plus tretinoin plus salicylic acid, indigo naturalise 1.4% ointment, and methotrexate gel. The effect size for the IAGI ranged from ‐0.56 (Sutton 2001; methotrexate gel) to ‐2.14 (Lin 2008; indigo naturalise 1.4% ointment).

In four treatments, the difference relative to placebo was not statistically significant: hexafluoro‐1,25‐dihydroxyvitamin D3, kukui nut oil, oleum horwathiensis, and platelet aggregation activating factor (PAF). No treatment was statistically significantly less effective than placebo.

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

This comparison included eight vitamin D analogue‐potent corticosteroid contrasts for body psoriasis (see Table 10). Eight studies with 2655 participants reported IAGI data for 6 of the 8 intervention‐comparator contrasts (see Analysis 7.1). Seven trials were between‐patient design, and one was a within‐patient study (Medansky 1996). Treatment duration ranged from three to eight weeks. Overall, there was no statistically significant difference between vitamin D analogues and potent corticosteroids: The SMD across all 6 treatments for IAGI was 0.17 (95% CI ‐0.04 to 0.37; I² statistic = 83.4%). In light of the high level of heterogeneity and inconsistency across treatments (Higgins 2011). We removed pooling. Vitamin D analogues performed statistically significantly better than one potent corticosteroid. This finding came from a single between‐patient study in which 99 participants contributed data (Bruce 1994). The SMD for calcipotriol against fluocinonide 0.05% ointment was ‐0.58 (95% CI ‐0.99 to ‐0.18; I² statistic = NA). Calcipotriol was statistically significantly less effective than both diflorasone diacetate 0.05% ointment (SMD 0.27; 95% CI 0.02 to 0.52) and betamethasone dipropionate (SMD 0.43; 95% CI 0.28 to 0.58; I² statistic = 50.3%). We found no statistically significant difference between calcipotriol and betamethasone valerate, calcitriol and betamethasone dipropionate, or calcitriol and betamethasone valerate.

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

This comparison included one vitamin D analogue, calcipotriol ointment, versus a very potent corticosteroid contrast, clobetasol propionate foam, for psoriasis of the body (see Table 11 and Analysis 8.1). One study with 42 participants reported IAGI data. Koo 2006 was a between‐patient study with a treatment duration of two weeks, which found no significant difference between the treatments (SMD 0.19; 95% CI ‐0.42 to 0.80).

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

This comparison considered vitamin D analogues‐steroid combination against potent or very potent corticosteroids for body psoriasis (see Analysis 9.1 and Table 12). The comparison included three contrasts: calcipotriol plus betamethasone dipropionate versus betamethasone dipropionate, calcipotriol plus betamethasone dipropionate versus clobetasol propionate, and calcipotriol plus clobetasol propionate versus clobetasol propionate.

Four between‐patient trials reported IAGI data for 1991 participants on 2 of the 3 intervention‐comparator contrasts. Treatment duration ranged between two and eight weeks. As these treatment comparisons were very different, we only pooled subtotals. In all but one trial (Fleming 2010 (H)), combination treatment was significantly more effective than corticosteroid alone. Three of the four trials compared calcipotriol/betamethasone dipropionate against betamethasone dipropionate (SMD ‐0.40; 95% CI ‐0.52 to ‐0.27; I² statistic = 41.8%), and one trial compared combined treatment with calcipotriol and clobetasol against clobetasol alone (SMD ‐0.69; 95% CI ‐1.22 to ‐0.15).

Analysis 10: Vitamin D alone or in combination versus dithranol

This comparison considered vitamin D analogues against dithranol (Analysis 10.1 and Table 13). We identified three intervention‐comparator contrasts: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. Five between‐patient trials reported IAGI data for 1108 participants on 2 of these 3 intervention‐comparator contrasts. Treatment duration ranged from 8 weeks to 12 weeks. There was some variation in the dithranol regimens employed by trials and in the baseline severity of trial participants. These factors may help explain the high level of heterogeneity found in the pooled results.

The SMD for the IAGI was ‐0.24 (95% CI ‐0.72 to 0.25; I² statistic = 93.0%). The presence of considerable heterogeneity means that the estimated average benefit should be treated with caution and pooling was therefore removed (Higgins 2011). Data from four trials contributed to the SMD for the calcipotriol versus dithranol: ‐0.43 (95% CI ‐0.85 to ‐0.01; I² statistic = 89.3%), indicating that calcipotriol was statistically significantly more effective than dithranol. Three of these four trials found a significant difference in favour of calcipotriol (Berth Jones 1992b; Christensen 1999; Wall 1998), but the trial by Van de Kerkhof 2006 found outpatient treatment with short contact dithranol to be significantly more effective than calcipotriol alone.

Data from one trial contributed to the SMD for the calcitriol versus dithranol: 0.51 (95% CI 0.13 to 0.88; I² statistic = NA), indicating that dithranol was statistically significantly more effective than calcitriol.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Our review identified three intervention‐comparator contrasts in this comparison: calcipotriol versus calcitriol, calcipotriol versus tacalcitol, and calcipotriol versus maxacalcitol for body psoriasis (see Analysis 11.1 and Table 14). Three trials involving 498 participants contributed IAGI data for all 3 intervention‐comparator contrasts (one trial for each contrast). Two trials were between‐patient, and one was within‐patient in design (Barker 1999 (H)). Treatment duration ranged from 8 to 12 weeks. The SMD for the IAGI was ‐0.06 (95% CI ‐0.51 to 0.38; I² statistic = 82.2%). The presence of substantial heterogeneity reflects differences in the findings from the two intervention‐comparator contrasts underlying this statistic. We found a statistically significant difference in favour of calcipotriol in the analysis against tacalcitol (SMD ‐0.47; 95% CI ‐0.73 to ‐0.21), but there was no significant difference relative to calcitriol (SMD 0.00; 95% CI ‐0.25 to 0.25) or between calcipotriol and maxacalcitol (SMD 0.43; 95% CI ‐0.12 to 0.98). In light of the inconsistency of results across treatments, we only pooled subtotals (Higgins 2011).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

Our review identified 12 intervention‐comparator contrasts for body psoriasis, involving 3 vitamin D analogues, 2 combination products, and 7 different corticosteroids (see Analysis 12.1 and Table 15 ). Eleven parallel‐group trials involving 4791 participants contributed IAGI data for 9 of these 12 intervention‐comparator contrasts. Treatment duration ranged from two to eight weeks. There were no IAGI data for three contrasts: calcipotriol versus calcipotriol plus diflucortolone valerate, calcipotriol versus calcipotriol plus fluocinonide acetonide, and calcitriol versus calcitriol plus diflucortolone valerate.

Overall, vitamin D alone appeared to be less effective than vitamin D plus corticosteroid: The SMD for the IAGI was 0.48 (95% CI 0.32 to 0.65; I² statistic = 86.9%). This finding applied to all but two of the intervention‐comparator contrasts: There was no statistically significant difference between twice‐daily calcipotriol and a regimen of calcipotriol (morning) plus betamethasone valerate (night time) (SMD 0.27; 95% CI ‐0.19 to 0.74) (Kragballe 1998b), or between once‐daily calcipotriol and once‐daily treatment with a combined product containing calcipotriol and hydrocortisone (SMD 0.14; 95% CI ‐0.06 to 0.33) (Ortonne 2010). The study by Ortonne 2010 also reported findings separately for the face (see Analysis 17.1).

In light of the observed heterogeneity between the intervention‐comparator contrasts, we only pooled subtotals (Higgins 2011).

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison summarises findings on complex regimens for body psoriasis, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments (see Analysis 13.1 and Table 16). We identified 12 intervention‐comparator contrasts, and IAGI data were available for 10 of these. Data from 2755 participants contributed to the IAGI analysis, based on findings from seven trials. Six of the trials were between‐patient (parallel‐group) in design, and one was a within‐patient study (Austad 1998). Trial duration varied between 6 and 12 weeks. As the interventions and comparators were highly variable and because two trials each contributed three pair‐wise contrasts, we did not pool the data.

Six intervention‐comparator contrasts for which IAGI data were available found a significant difference between regimens. A two‐week regimen with clobetasol propionate followed by four weeks' treatment with calcipotriol was more effective than six weeks of monotherapy with calcipotriol (SMD 0.60; 95% CI 0.18 to 1.02). All treatments were applied twice‐daily in this within‐patient study (Austad 1998).

The remaining five more effective regimens all involved once‐daily treatment with a combined product containing calcipotriol and betamethasone dipropionate, and the same study (White 2006 (P)) reported three contrasts. This study involved an initial 'treatment' phase consisting of 4 weeks' treatment with the combined product, followed by an 8‐week maintenance phase. In one maintenance phase, participants received placebo ointment for 8 weeks. This was significantly less effective than maintenance with once‐daily calcipotriol (SMD 0.27; 95% CI 0.12 to 0.41), and it was also less effective than maintenance with once‐daily calcipotriol on weekdays and the combined product at the weekend (SMD 0.51; 95% CI 0.37 to 0.66). When these two comparator regimens were compared directly (White 2006 (H)), the alternating regimen using weekday/weekend treatments during the maintenance phase was significantly more effective than once‐daily calcipotriol for maintenance (SMD 0.26; 95% CI 0.11 to 0.40). Kragballe 2004 compared 3 12‐week treatment regimens, 2 of which were 'complex'. When we compared these complex regimens, eight weeks' once‐daily combination treatment with calcipotriol and betamethasone dipropionate, followed by four weeks of once‐daily calcipotriol was significantly less effective than a regimen consisting of four weeks' once‐daily combination treatment, followed by eight weeks of once‐daily calcipotriol on weekdays and the combined product at weekends (SMD 0.24; 95% CI 0.08 to 0.40). A separate trial compared 8 weeks with tacalcitol to combination ointment for 4 weeks, followed by calcipotriol for 4 weeks (Ortonne 2004). Participants applied all treatments once daily (at night time). Monotherapy with tacalcitol was significantly less effective than the complex regimen (SMD 0.54; 95% CI 0.36 to 0.72).

In 4 of the 10 intervention‐comparator contrasts for which IAGI data were available, there was no significant difference in effect. The study by Kragballe 2004 (see paragraph above) found no difference between twice‐daily calcipotriol and the two comparator complex regimens. One regimen consisted of eight weeks' once‐daily treatment with a combined product containing calcipotriol and betamethasone dipropionate, followed by four weeks of once‐daily calcipotriol. Relative to calcipotriol monotherapy, the SMD for the IAGI was ‐0.12 (95% CI ‐0.29 to 0.04). When compared with a regimen consisting of four weeks' once‐daily treatment with a combined product containing calcipotriol and betamethasone dipropionate, followed by eight weeks of once‐daily calcipotriol on weekdays and the combined product at weekends, twice‐daily calcipotriol appeared to be slightly less effective, although the difference was not statistically significant (SMD 0.13; 95% CI ‐0.04 to 0.29).

Yang 2009 compared a six‐week course of twice‐daily calcipotriol with a complex routine using halometasone and calcipotriol. For two weeks participants applied halometasone in the morning and calcipotriol at night; over the next two weeks, halometasone was applied twice daily at weekends and calcipotriol twice daily during the week. For the final two weeks, treatment reverted to twice‐daily calcipotriol. Although there was a trend in favour of the complex regimen (SMD 0.41; 95% CI ‐0.05 to 0.86), the difference was not statistically significant.

Lahfa 2003 compared two different vitamin D products combined with a very potent corticosteroid. In the initial phase, participants applied dual therapy (clobetasol propionate in the mornings and the vitamin D analogue (calcipotriol or calcitriol) at night) until achieving clearance or marked improvement or until 4 weeks had elapsed. Participants then switched to a monotherapy maintenance phase with the vitamin D product for the remainder of the 12‐week study period. The IAGI results showed the 2 treatment regimens to be equivalent: ‐0.19 (95% CI ‐0.54 to 0.16).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

This comparison included active‐controlled studies of psoriasis of the body that were at least 24 weeks in duration (see Analysis 14.1 and Table 17). One longer‐term placebo‐controlled trial of the body (Katz 1991a) and two long‐term scalp trials (Luger 2008; Poulin 2010) did not meet the inclusion criteria for this comparison, and they are evaluated elsewhere (Analysis 2, Analysis 19, and Analysis 18, respectively).

One trial was included in this comparison; Kragballe 2006 was a between‐patient trial that compared three 52‐week regimens with all treatments used once daily:

• first, combination treatment with calcipotriol and betamethasone dipropionate;

• second, alternating treatment with combination therapy for 4 weeks, then calcipotriol for 4 weeks; and

• third, combination therapy for 4 weeks, then calcipotriol for 48 weeks.

In total, 297 participants contributed IGA data to the analysis. Data were unsuitable for pooling because the same participants contributed to more than one analysis.

According to the investigators' assessment, there was no significant difference between the three long‐term regimens. One year's combination therapy was not significantly better than either the alternating regimen (SMD ‐0.09; 95% CI ‐0.36 to 0.18) or the regimen of treatment with 4 weeks of combination therapy followed by 48 weeks of calcipotriol (SMD ‐0.18; 95% CI ‐0.47 to 0.10). In both these comparisons, combination therapy achieved a larger absolute benefit, but the difference was not statistically significant. When alternating therapy was compared with the regimen of 4 weeks' combination therapy followed by 48 weeks of calcipotriol, the SMD for the IAGI also indicated the two regimens were not statistically significantly different: ‐0.09 (95% CI ‐0.37 to 0.19).

Analysis 15: Vitamin D analogues versus other treatment

This comparison incorporated all other vitamin D head‐to‐head comparisons of treatments for psoriasis of the body (excluding inverse psoriasis) that had not already been included (see Analysis 15.1 and Table 18). We included 12 intervention‐comparator contrasts, with IAGI data available for 6 of these contrasts. Eight between‐patient trials and 2 within‐patient trials provided data for 1386 participants. Trial duration ranged between 4 and 12 weeks. In light of the pharmacological diversity amongst the comparators, we only pooled data within subgroups.

According to the investigator's global assessment, twice‐daily calcipotriol was significantly more effective than coal tar polytherapy (SMD ‐0.59; 95% CI ‐0.87 to ‐0.31; I² statistic = 0%). Once‐daily vitamin D was significantly less effective than a twice‐daily application (SMD ‐0.24; 95% CI ‐0.38 to ‐0.09; I² statistic = 0%). This contrast included a comparison of once‐ versus twice‐daily dosing of calcipotriol (SMD ‐0.27; 95% CI ‐0.48 to ‐0.06) and a dosing comparison combination treatment with calcipotriol and betamethasone dipropionate (SMD ‐0.20; 95% CI ‐0.41 to 0.00).

In the remaining four intervention‐comparator contrasts, no significant difference was found between twice‐daily calcipotriol and the comparators. This finding held for the comparison against coal tar monotherapy (SMD ‐0.53; 95% CI ‐1.74 to 0.68; I² statistic = 91.1%), against betamethasone dipropionate ointment and salicylic acid (SMD ‐0.06; 95% CI ‐0.33 to 0.22; I² statistic = NA), against tacrolimus ointment (SMD ‐0.22; 95% CI ‐0.60 to 0.16; I² statistic = NA), and against vitamin B12 cream (SMD ‐0.55; 95% CI ‐1.33 to 0.24; I² statistic = NA).

Two of the three trials comparing calcipotriol with coal tar monotherapy found a significant difference in favour of calcipotriol; the other trial found a significant difference in favour of coal tar (Alora‐Palli 2010). This apparent contradiction may reflect different formulations of coal tar, treatment durations, or different baseline disease severity.

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of topical treatments for inverse or facial psoriasis (see Analysis 16.1 and Table 19). Evidence on four treatments was found in this comparison: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus. We only found one trial that evaluated tacrolimus ointment (Lebwohl 2004), but the study did not report any effectiveness data suitable for this review. However, the study did contribute data on adverse events and withdrawal rates.

IAGI outcome data were available for one of the four topical treatments for inverse psoriasis. One 8‐week between‐patient study with 47 participants (Gribetz 2004) reported placebo‐controlled data on pimecrolimus 1% cream, which demonstrated a statistically significant difference in favour of twice‐daily pimecrolimus (SMD ‐1.07; 95% CI ‐1.69 to ‐0.45).

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for inverse psoriasis, which compared vitamin D with an active control (see Analysis 17.1 and Table 20). We identified five intervention‐comparator contrasts. Four treatments were compared with calcipotriol: once‐daily betamethasone valerate, combined treatment with calcipotriol and hydrocortisone, calcitriol, and pimecrolimus. Calcitriol was compared with tacrolimus.

Two between‐patient studies contributed IAGI data from 457 study participants on 2 of the 5 intervention‐comparator contrasts. Trial duration ranged from six to eight weeks. When applied once daily to inverse psoriasis, calcipotriol was significantly less effective than combined treatment with calcipotriol and hydrocortisone (SMD 0.30; 95% CI 0.11 to 0.50) (Ortonne 2010). However, there was no significant difference in effect between twice daily treatment with calcitriol and tacrolimus (SMD 0.42; 95% CI ‐0.15 to 0.98) (Liao 2007).

Analysis 18: Scalp psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of treatments for scalp psoriasis (see Analysis 18.1 and Table 21). We included evidence on 11 treatments in this comparison, with IAGI data available for 9 treatments. Ten trials, 9 of which were between‐patient studies, contributed data from 2472 participants. One study was a within‐patient trial (Lepaw 1978). Trial duration ranged between two and eight weeks. IAGI data were not available for betamethasone valerate or ciclopirox olamine shampoo.

Eight treatments for scalp psoriasis that were assessed using the IAGI scale were significantly more effective than placebo. The least effective treatment was calcipotriol (SMD ‐0.72; 95% CI ‐1.28 to ‐0.16; I² statistic = 69.2%), and the most effective treatment was clobetasol propionate (SMD ‐1.73; 95% CI ‐1.99 to ‐1.48; I² statistic = 14.3%). Combination treatment with calcipotriol and betamethasone dipropionate was also effective (SMD ‐0.97; 95% CI ‐1.61 to ‐0.32; I² statistic = 90.2%). Data from the three very potent corticosteroids were pooled (amcinonide, clobetasol propionate, halcinonide). The SMD across these treatments was ‐1.57 (95% CI ‐1.85 to ‐1.28; I² statistic = 49.7%). Only salicylic acid was not significantly more effective than placebo (SMD ‐0.86; 95% CI ‐1.79 to 0.06).

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.1 and Table 22). We identified six intervention‐comparator contrasts, and IAGI data were available for five of these contrasts. All studies were parallel‐group in design (between‐patient). Ten studies contributed IAGI data from 5175 participants, and trial duration ranged from 4 to 52 weeks.

Vitamin D for scalp psoriasis was significantly less effective than potent steroids, either alone or in combination with vitamin D. Specifically, calcipotriol was less effective than three comparator treatments: betamethasone dipropionate (SMD 0.48; 95% CI 0.32 to 0.64; I² statistic = 60.4%), betamethasone valerate (SMD 0.37; 95% CI 0.20 to 0.55; I² statistic = 0%), and combination treatment with calcipotriol and betamethasone dipropionate (SMD 0.64; 95% CI 0.44 to 0.84; I² statistic = 82.3%). Combination treatment (calcipotriol/betamethasone dipropionate) was significantly more effective than betamethasone dipropionate alone (SMD ‐0.18; 95% CI ‐0.26 to ‐0.10; I² statistic = 0%). The efficacy of calcipotriol and coal tar polytherapy was similar: SMD ‐0.24 (95% CI ‐0.73 to 0.25; I² statistic = 91.1%).

Analysis 1: Vitamin D analogues versus placebo

Our review included eight vitamin D analogues for body psoriasis in this comparison (see Analysis 1.2 and Table 5). Nineteen studies reported TSS data with 2647 participants contributing data. Nine trials were between‐patient design, and 10 were within‐patient studies. Treatment duration ranged from 4 weeks to 12 weeks. The average effect size across all 8 treatments for TSS was ‐1.04; (95% CI ‐1.33 to ‐0.74; I² statistic = 93.0%), but two of these treatments were not significantly more effective than placebo. The high level of heterogeneity means that the estimated average benefit should be treated with caution; there was considerable variation between effective treatments, with the TSS ranging from ‐0.46 (becocalcidiol twice daily) to ‐2.15 (paricalcitol once daily). Therefore, we removed pooling across subgroups (Higgins 2011).

Four studies contributed data for the SMD for the TSS of calcitriol (SMD ‐1.22; 95% CI ‐2.38 to ‐0.07), but there was considerable heterogeneity within the subgroup (I² statistic = 98.3%). One of the studies (Van de Kerkhof 1989) found no statistically significant difference between calcitriol and placebo (SMD ‐0.06; 95% CI ‐0.94 to 0.81), whereas the study by Perez 1996 found a large and statistically significant difference (‐4.03; 95% CI ‐4.56 to ‐3.50). We considered this finding in more detail in the 'combined end point' section (e) below.

Analysis 2: Corticosteroid (potent) versus placebo

Our review included 10 potent corticosteroids for body psoriasis in this comparison (see Analysis 2.2 and Table 6). Seven studies reported TSS data contributed by 553 participants on 8 of these 10 treatments. Six trials were between‐patient studies, and there was one within‐patient design (Ormerod 1997). Treatment duration ranged from 2 to 12 weeks. The average effect size across all 8 treatments for TSS was ‐0.77 (95% CI ‐1.01 to ‐0.52; I² statistic = 46.7%). We found all treatments, except for diflorasone diacetate, to be statistically significantly superior to placebo: SMD ‐0.32; 95% CI ‐0.73 to 0.09).

Analysis 3: Corticosteroid (very potent) versus placebo

Of the three very potent corticosteroids for body psoriasis included in this comparison, TSS data were available only for clobetasol propionate (see Analysis 3.2 and Table 7). Three studies, all of which were between‐patient trials, reported TSS data for 545 participants. Treatment duration ranged from two to four weeks. The SMD for the TSS was ‐1.35 (95% CI ‐1.80 to ‐0.89; I² statistic = 75.3%). In Analysis 18.2, we reported TSS data on the use of very potent steroids on the scalp.

Analysis 4: Dithranol versus placebo

This comparison considered dithranol against placebo (see Analysis 4.2 and Table 23). Three within‐patient trials reported TSS data for 47 participants. Treatment duration ranged from three to eight weeks. The SMD for the TSS was ‐1.06 (95% CI ‐1.66 to ‐0.46; I² statistic = 37.4%).

Analysis 5: Vitamin D combination products versus placebo

Our review did not identify any trial reporting TSS data for this comparison.

Analysis 6: Other treatment versus placebo

This comparison comprised all other treatments for body psoriasis that were not included in the first five comparisons. None of the studies assessed the same treatment, which means that findings should be interpreted with caution.

In total, we included 26 treatments in this analysis (see Analysis 6.2 and Table 9). Seventeen studies with 907 participants reported TSS data on 17 of these 26 treatments. Five trials were between‐patient design, and 12 were within‐patient studies. Treatment duration ranged from 3 to 12 weeks.

Ten treatments performed significantly better than placebo: anti‐IL‐8 monoclonal antibody cream, calcipotriene 0.005% ointment + nicotinamide, fish oil plus occlusion, hexafluoro‐1,25‐dihydroxyvitamin D3, indigo naturalise 1.4% ointment, methotrexate gel, mycophenolic acid ointment, oleum horwathiensis, PTH (1‐34) in Novasome cream®, and tazarotene. The effect size for the TSS ranged from ‐0.48 (Levine 2010 (P); calcipotriene 0.005% ointment + nicotinamide) to ‐2.31 (Holick 2003; PTH (1‐34) in Novasome cream®).

In seven treatments, the difference relative to placebo was not statistically significant: kukui nut oil, NG‐monomethyl‐L‐arginine (L‐NMMA) cream, nicotinamide 1.4%, polymyxin B cream, topical sirolimus, topical tacrolimus, and tar.

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

There were eight vitamin D analogue‐potent corticosteroid comparisons for body psoriasis (see Analysis 7.2 and Table 10). Six studies with 891 participants reported TSS data for 5 of the 8 intervention‐comparator contrasts. Two trials were between‐patient design, and four were within‐patient studies. Treatment duration ranged from three to six weeks. The SMD across all 5 treatments for TSS indicated that there was no significant difference between the vitamin D derivatives and potent corticosteroid: SMD 0.11 (95% CI ‐0.22 to 0.44; I² statistic = 86.7%). However, there was considerable variation between the individual contrasts underlying the pooled effect. In light of this heterogeneity, we removed pooling (Higgins 2011).

In two of the five vitamin D‐potent corticosteroid comparisons, the vitamin D analogue performed statistically significantly better than the potent corticosteroid: The SMD for calcipotriol against fluocinonide 0.05% ointment was ‐0.50 (95% CI ‐0.92 to ‐0.07; I² statistic = NA) (Bruce 1994). Similarly, the comparison of calcipotriol against betamethasone valerate showed a significant difference in favour of calcipotriol (SMD ‐0.26; 95% CI ‐0.41 to ‐0.11), a finding also based on a single study (Kragballe 1991a).

In three comparisons, the vitamin D analogue was statistically significantly less effective than the potent corticosteroid: calcipotriol versus diflorasone diacetate (SMD 0.40; 95% CI 0.15 to 0.65), calcitriol versus betamethasone dipropionate (SMD 0.27; 95% CI 0.02 to 0.51), and tacalcitol versus betamethasone valerate (SMD 0.41; 95% CI 0.09 to 0.74).

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

Our review did not identify any study of psoriasis of the body that compared vitamin D analogues against very potent corticosteroids and that reported TSS data.

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

This comparison considered vitamin D analogues‐steroid combination against potent or very potent corticosteroid for body psoriasis (see Analysis 9.2 and Table 12). One four‐week between‐patient trial reported TSS data for 122 participants with moderate to severe psoriasis (Menter 2009). The combined treatment with calcipotriol and betamethasone dipropionate was found to be significantly less effective than clobetasol propionate spray alone (0.45; 95% CI 0.09 to 0.81).

Analysis 10: Vitamin D alone or in combination versus dithranol

This comparison considered vitamin D analogues against dithranol (see Analysis 10.2 and Table 13). We identified three intervention‐comparator contrasts: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. Three between‐patient trials and 1 within‐patient trial (Grattan 1997 (H)) reported TSS data for 386 participants. Treatment duration ranged from four weeks to eight weeks. There was some variation in the dithranol regimens employed by trials and in the baseline severity of trial participants. These factors may help explain the substantial heterogeneity found in the pooled results (Higgins 2011).

The SMD for the TSS was ‐0.27 (95% CI ‐0.73 to 0.20; I² statistic = 80.6%). Data from two trials contributed to the SMD for the calcipotriol versus dithranol: ‐0.54 (95% CI ‐1.16 to 0.08; I² statistic = 71.2%) (Christensen 1999; Grattan 1997 (H)). Data from one trial contributed to the SMD for the calcitriol versus dithranol: 0.13 (95% CI ‐0.24 to 0.50; I² statistic = NA) (Hutchinson 2000). Data from one trial contributed to the SMD for the tacalcitol versus dithranol: ‐0.18 (95% CI ‐0.60 to 0.25; I² statistic = NA) (Farkas 1999). Therefore, neither the summary statistic for the TSS nor the pooled data for individual intervention‐comparator contrasts provided evidence of a statistically significant advantage of a vitamin D analogue over dithranol or vice versa.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Our review identified three intervention‐comparator contrasts for body psoriasis in this comparison: calcipotriol versus calcitriol, calcipotriol versus tacalcitol, and calcipotriol versus maxacalcitol (see Analysis 11.2 and Table 14). Three trials involving 563 participants contributed TSS data for all 3 of these intervention‐comparator contrasts. Two trials were between‐patient, and one was within‐patient in design (Barker 1999 (H)). Treatment duration ranged from 8 to 12 weeks. The SMD for the TSS indicated that there was a statistically significant difference in favour of calcipotriol: SMD ‐0.31; 95% CI ‐0.55 to ‐0.06; I² statistic = 46.9%. When assessed by the TSS, calcipotriol was statistically significantly more effective than calcitriol (SMD ‐0.32; 95% CI ‐0.57 to ‐0.07). This finding contrasts with the IAGI assessment from the same study, which found no significant difference (see Ji 2008;Analysis 11.1). Calcipotriol was also significantly more effective than tacalcitol (SMD ‐0.45; 95% CI ‐0.68 to ‐0.22) and similar in efficacy relative to maxacalcitol (SMD 0.13; 95% CI ‐0.41 to 0.68).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

Our review identified 12 intervention‐comparator contrasts for body psoriasis, involving 3 vitamin D analogues, 2 combination products, and 7 different corticosteroids (see Analysis 12.1 and Table 15). One 4‐week parallel‐group trial contributed TSS data from 301 participants for 1 of these 12 intervention‐comparator contrasts (Huang 2009). Twice‐daily calcipotriol was found to be statistically significantly less effective than once‐daily treatment with a combined product containing calcipotriol and betamethasone dipropionate (SMD 0.25; 95% CI 0.03 to 0.48).

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison summarises findings on complex regimens for body psoriasis, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments (see Analysis 13.2 and Table 16). We identified 12 intervention‐comparator contrasts and TSS data were available for one of these. Data from 1 6‐week within‐patient trial with 46 participants (Austad 1998) contributed to the TSS analysis. Austad 1998 compared six weeks of twice‐daily calcipotriol with a regimen of two weeks' treatment with clobetasol propionate followed by four weeks with calcipotriol. The complex regimen was significantly more effective than monotherapy with calcipotriol (SMD 0.63; 95% CI 0.21 to 1.05).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

We did not identify any relevant study that provided TSS data for this comparison.

Analysis 15: Vitamin D analogues versus other treatment

This comparison incorporated all other vitamin D head‐to‐head comparisons of treatments for psoriasis of the body (excluding inverse psoriasis) that had not already been included (see Analysis 15.2 and Table 18). We included 12 intervention‐comparator contrasts, with TSS data available for 6 of these contrasts. Five between‐patient trials and 2 within‐patient trials provided data from 898 participants. Trial duration ranged between 6 and 12 weeks. In light of the pharmacological diversity of the comparators, we only pooled data within subgroups.

According to the TSS assessment, twice‐daily calcipotriol was significantly more effective than coal tar polytherapy (SMD ‐0.51; 95% CI ‐0.86 to ‐0.16; I² statistic = NA). In the remaining five intervention‐comparator contrasts, we found no significant difference between twice‐daily calcipotriol and the comparators.

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of topical treatments for inverse or facial psoriasis (see Analysis 16.2 and Table 19). Evidence on four treatments was found in this comparison: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus. We found only one placebo‐controlled trial that evaluated tacrolimus ointment (Lebwohl 2004), but the study did not report any effectiveness data suitable for this review. However, the study did contribute data on adverse events and withdrawal rates.

TSS data were available for one of the four topical treatments for inverse psoriasis. One 8‐week between‐patient study contributed data on pimecrolimus 1% cream from 57 participants (Gribetz 2004), 10 participants more than those with an investigator's assessment (see Analysis 16.1). Findings from the TSS were consistent with those of the IAGI: the SMD for the TSS also found a statistically significant difference in favour of twice‐daily pimecrolimus (SMD ‐1.37; 95% CI ‐1.95 to ‐0.79).

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for inverse psoriasis, which compared vitamin D with an active control (see Analysis 17.2 and Table 20). We identified five intervention‐comparator contrasts. Four treatments were compared with calcipotriol: once‐daily betamethasone valerate, combined treatment with calcipotriol and hydrocortisone, calcitriol, and pimecrolimus. Calcitriol was compared with tacrolimus.

Two 6‐week studies contributed TSS data from 124 study participants on 2 of the 5 intervention‐comparator contrasts. Calcipotriol was significantly less effective than calcitriol (SMD 0.61; 95% CI 0.28 to 0.94) (Ortonne 2003). In this within‐patient study, participants applied treatments twice daily to 'sensitive areas', including the face, hairline, retro‐auricular, and flexural areas. When they applied treatments to the facial and genitofemoral areas, there was no significant difference in effect between twice‐daily treatment with calcitriol and tacrolimus (SMD 0.29; 95% CI ‐0.27 to 0.85) (Liao 2007). In this between‐patient study, participants applied both treatments twice daily.

Analysis 18: Scalp psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of treatments for scalp psoriasis (see Analysis 18.2 and Table 21). We included evidence on 11 treatments in this comparison, with TSS data available for 10 treatments. Twelve between‐patient trials contributed data from 2897 participants. Trial duration ranged between two and eight weeks. TSS data were not available for halcinonide (classed as a very potent corticosteroid).

Eight of the 10 treatments for scalp psoriasis that were assessed using the TSS scale were significantly more effective than placebo. The least effective treatment was calcipotriol (SMD ‐0.44; 95% CI ‐0.64 to ‐0.25; I² statistic = 0%), and the most effective were the 2 very potent steroids, amcinonide (SMD ‐1.58; 95% CI ‐1.98 to ‐1.18) and clobetasol propionate (SMD ‐1.53; 95% CI ‐1.77 to ‐1.28; I² statistic = 46.3%). Combination treatment with calcipotriol and betamethasone dipropionate was also effective (SMD ‐0.92; 95% CI ‐1.42 to ‐0.43; I² statistic = 83.2%).

We pooled data from the two potent corticosteroids, betamethasone valerate and betamethasone dipropionate (SMD ‐1.13; 95% CI ‐1.44 to ‐0.81; I² statistic = 52.2%). We also pooled data from the two very potent corticosteroids, amcinonide and clobetasol propionate (SMD ‐1.55; 95% CI ‐1.73 to ‐1.37; I² statistic = 19.6%).

Two treatments were not significantly different to placebo: ciclopirox olamine shampoo (SMD ‐0.07; 95% CI ‐0.82 to 0.68) and salicylic acid (SMD ‐0.57; 95% CI ‐1.47 to 0.32).

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.2 and Table 22).

We identified six intervention‐comparator contrasts, and TSS data were available for all six of these contrasts. All studies were parallel‐group in design (between‐patient). Eleven studies contributed TSS data from 4877 participants, and trial duration ranged from 4 to 8 weeks.

Based on Total Severity Scores, calcipotriol was significantly less effective than betamethasone dipropionate (SMD 0.45; 95% CI 0.28 to 0.63; I² statistic = 66.3%), clobetasol propionate (SMD 0.37; 95% CI 0.05 to 0.69; I² statistic = NA), and combination treatment with calcipotriol and betamethasone dipropionate (SMD 0.70; 95% CI 0.56 to 0.84; I² statistic = 49.6%). There was no statistically significant difference between calcipotriol and betamethasone valerate (SMD 0.09; 95% CI ‐0.09 to 0.27; I² statistic = 0%). Combination treatment (calcipotriol/betamethasone dipropionate) was significantly more effective than betamethasone dipropionate alone (SMD ‐0.19; 95% CI ‐0.27 to ‐0.11; I² statistic = 0%). The efficacy of calcipotriol and coal tar polytherapy was not significantly different (SMD ‐0.30; 95% CI ‐0.84 to 0.24; I² statistic = 93.1%).

Analysis 1: Vitamin D analogues versus placebo

Our review included eight vitamin D analogues for body psoriasis in this comparison (see Analysis 1.3 and Table 5). Nine studies reported PASI data, with 2357 participants contributing data on 2 (calcipotriol and tacalcitol) of these 8 treatments. Eight trials were between‐patient design, and one was a within‐patient study (Dubertret 1992). Treatment duration ranged from three weeks to eight weeks. The average effect size across the 2 treatments for PASI was SMD ‐0.58 (95% CI ‐0.71 to ‐0.45; I² statistic = 42.3%).

Analysis 2: Corticosteroid (potent) versus placebo

Our review included 10 potent corticosteroids for body psoriasis in this comparison (see Analysis 2.3 and Table 6). Three between‐patient studies reported PASI data from 1158 participants on once‐daily or twice‐daily doses of betamethasone dipropionate (2 of the 10 potent steroid regimens considered in this group). Trial duration ranged between four and eight weeks. The average PASI effect size across both application frequencies was ‐0.97 (95% CI ‐1.31 to ‐0.62; I² statistic = 79.6%).

Analysis 3: Corticosteroid (very potent) versus placebo

Our review did not identify any study that compared very potent corticosteroids against placebo and also reported PASI data. This may be because whole‐body application of very potent corticosteroids is not recommended, so a whole‐body assessment measure like the PASI is inappropriate.

Analysis 4: Dithranol versus placebo

Our review did not identify any study that compared dithranol against placebo and also reported PASI data.

Analysis 5: Vitamin D combination products versus placebo

This comparison included treatment with combined calcipotriol and betamethasone dipropionate used either once or twice daily on the body (see Analysis 5.3 and Table 8). Five parallel‐group studies with 2056 participants contributed PASI data on both dosing options. Treatment duration ranged from four to eight weeks. The PASI SMD across treatments was ‐1.24 (95% CI ‐1.53 to ‐0.95; I² statistic = 87.6%). Although twice‐daily combination treatment (SMD ‐1.41; 95% CI ‐1.86 to ‐0.97) achieved a larger effect than once‐daily treatment (SMD ‐1.14; 95% CI ‐1.57 to ‐0.70), the difference was not statistically significant at the 5% level. This finding contrasts with the assessment of the relative benefit of once‐ versus twice‐daily dosing using the IAGI metric (see Analysis 5.1).

Analysis 6: Other treatment versus placebo

This comparison comprised all other treatments for body psoriasis that were not included in the first five comparisons; therefore, we removed pooling. None of the studies assessed the same treatment, which means that findings should be interpreted with caution.

In total, we included 26 treatments in this analysis (see Analysis 6.3 and Table 9). Nine studies with 529 participants reported PASI data on 9 of these 26 treatments. Eight trials were between‐patient studies, and one was within‐patient in design (Vali 2005). Treatment duration ranged from 2 to 12 weeks.

Six treatments performed statistically significantly better than placebo: aloe vera extract, betamethasone 17‐valerate 21‐acetate plus tretinoin plus salicylic acid, a herbal skin care product, Mahonia aquifolium, methotrexate gel, and theophylline ointment. Effects sizes ranged from ‐0.54 (95% CI ‐0.99 to ‐0.10) for the betamethasone 17‐valerate 21 product (Santoianni 2001) to ‐2.96 (95% CI ‐4.19 to ‐1.74) for the herbal skin care treatment (Maier 2004).

In three treatments, we found no statistically significant difference relative to placebo in the PASI assessments. These comprised topical caffeine (Vali 2005), dead sea salts emollient lotion (Cheesbrough 1992), and kukui nut oil (Brown 2005).

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

Our review identified eight vitamin D analogue‐potent corticosteroid comparisons for body psoriasis (see Analysis 7.3 and Table 10). Nine studies with 3185 participants reported PASI data for 4 of the 8 intervention‐comparator contrasts. Seven trials were between‐patient design, and two were within‐patient studies. Treatment duration ranged from four to eight weeks. The PASI SMD across all four intervention‐comparator contrasts indicated that there was no statistically significant difference between the vitamin D derivatives and potent corticosteroid (SMD 0.12; 95% CI ‐0.07 to 0.32; I² statistic = 86.2%).

In one vitamin D analogue versus potent corticosteroid comparison, the vitamin D analogue performed statistically significantly better than the potent corticosteroid: four studies contributed data to analysis of calcipotriol versus betamethasone valerate (SMD ‐0.12; 95% CI ‐0.22 to ‐0.02; I² statistic = 0%).

In two intervention‐comparator contrasts, the vitamin D analogue was statistically significantly less effective than the potent corticosteroid: Betamethasone dipropionate was more effective than both calcipotriol (SMD 0.36; 95% CI 0.22 to 0.51; I² statistic = 49.6%) and calcitriol (SMD 0.39; 95% CI 0.14 to 0.63; I² statistic = NA).

We found no statistically significant difference between calcipotriol and desoxymetasone.

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

This comparison considered vitamin D analogues against very potent corticosteroids for treating psoriasis of the body (see Analysis 8.3 and Table 11). We found data on one intervention‐comparator contrast: calcipotriol versus clobetasol propionate. One between‐patient trial reported PASI data for 40 participants. This trial had a treatment duration of six weeks (Landi 1993). The SMD for the PASI indicated that there was no statistically significant difference between the very potent corticosteroid and the vitamin D analogue: SMD ‐0.32 (95% CI ‐0.95 to 0.30; I² statistic = NA).

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

This comparison considered vitamin D analogue‐steroid combination against potent or very potent corticosteroid for body psoriasis (Table 12). Three between‐patient trials reported PASI data for 1876 participants (see Analysis 9.3). All 3 trials compared combination treatment with calcipotriol and betamethasone dipropionate with betamethasone dipropionate as monotherapy, but varied in their treatment duration (4 to 8 weeks). The SMD for the PASI was ‐0.44 (95% CI ‐0.55 to ‐0.33; I² statistic = 22.4%), indicating a significantly greater effect for combination treatment.

Analysis 10: Vitamin D alone or in combination versus dithranol

This comparison considered vitamin D analogues against dithranol (see Analysis 10.3 and Table 13). We identified three intervention‐comparator contrasts: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. Five between‐patient trials reported PASI data for 796 participants. Treatment duration ranged from 8 to 12 weeks. There was some variation in the dithranol regimens employed by trials and in the baseline severity of trial participants. These factors may explain the considerable level of heterogeneity found in the pooled results (Higgins 2011).

The SMD for the PASI was 0.36 (95% CI ‐0.33 to 1.04; I² statistic = 94.5%). Data from three trials contributed to the SMD for the calcipotriol versus dithranol: 0.73 (95% CI ‐0.55 to 2.00; I² statistic = 97.2%) (Berth Jones 1992b; Monastirli 2000; Van de Kerkhof 2006). One of these three trials found a large and statistically significant difference in favour of calcipotriol (Berth Jones 1992b); Monastirli 2000 found a significant difference in favour of dithranol, and the trial by Van de Kerkhof 2006 found no difference between the two treatments. In the light of this heterogeneity, we removed all pooling from this comparison.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Our review identified three intervention‐comparator contrasts for body psoriasis in this comparison: calcipotriol versus calcitriol, calcipotriol versus tacalcitol, and calcipotriol versus maxacalcitol (see Analysis 11.3 and Table 14). One between‐patient trial involving 15 participants contributed PASI data for the comparison of calcipotriol and calcitriol (Bourke 1997). Treatment duration was eight weeks. Although there was a trend towards a greater effect for calcipotriol, the SMD for the PASI was not statistically significant: ‐1.11 (95% CI ‐2.22 to 0.01; I² statistic = NA).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

Our review identified 12 intervention‐comparator contrasts, involving 3 vitamin D analogues, 2 combination products, and 7 different corticosteroids (see Analysis 12.3 and Table 15). Sixteen trials involving 5703 participants contributed PASI data for 11 of these 12 intervention‐comparator contrasts. Fifteen trials were between‐patient, and 1 was within‐patient in design (Salmhofer 2000). Treatment duration ranged from 2 to 12 weeks. We did not identify any PASI data for the comparison of calcipotriol against clobetasol propionate then calcipotriol. Overall, vitamin D plus corticosteroid appeared to be more effective than vitamin D alone: The SMD for the PASI was 0.47 (95% CI 0.34 to 0.59; I² statistic = 82.3%). However, there was considerable variation between the findings of the intervention‐comparator contrasts. In light of the observed heterogeneity, we only pooled subtotals (Higgins 2011).

In six of the intervention‐comparator contrasts, the analysis of the PASI measure found no significant difference between the treatments. This was the case for twice‐daily calcipotriol compared with the following:

• a regimen with calcipotriol applied in the morning and the night‐time application of betamethasone valerate (Kragballe 1998b; Ruzicka 1998);

• clobetasone butyrate (Kragballe 1998b); or

• diflucortolone valerate (Salmhofer 2000).

Once‐daily calcipotriol was no more effective than combined treatment with fluocinonide acetonide (Wozel 2001) or combined with hydrocortisone (Ortonne 2010). Lastly, no significant difference was found between twice‐daily calcitriol and combined treatment with diflucortolone valerate (mornings) and calcitriol (night time) (Lee 2007).

In contrast, combined treatment with calcipotriol and betamethasone dipropionate was significantly more effective than twice‐daily calcipotriol alone. This finding held for regimens involving night‐time applications of betamethasone dipropionate (SMD 0.46; 95% CI 0.10 to 0.82; I² statistic = NA) and for treatment with a combined product, whether applied once daily (SMD 0.52; 95% CI 0.38 to 0.67; I² statistic = 32.3%) or twice daily (SMD 0.64; 95% CI 0.46 to 0.83; I² statistic = 73.6%). Once‐daily treatment with the combined calcipotriol/betamethasone dipropionate product was also significantly more effective than once‐daily calcipotriol (SMD 0.67; 95% CI 0.23 to 1.11; I² statistic = 87.4%) as well as once‐daily tacalcitol (SMD 0.47; 95% CI 0.25 to 0.69; I² statistic = NA).

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison summarises findings on complex regimens for body psoriasis, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments (see Analysis 13.3 and Table 16). PASI data were available for 11 of the 12 intervention‐comparator contrasts identified. Data from 2991 participants contributed to the PASI analysis, based on findings from 8 between‐patient trials. Trial duration varied between 2 and 12 weeks. Because the interventions and comparators were highly variable and two trials each contributed three pair‐wise contrasts, we did not pool the data.

Six intervention‐comparator contrasts for which PASI data were available found a significant difference between regimens. A six‐week course of calcipotriol was compared with two different complex regimens. Monotherapy with calcipotriol was significantly less effective than two weeks of treatment with calcipotriol and fluocinonide acetonide followed by four weeks of calcipotriol (SMD 0.66; 95% CI 0.01 to 1.32) (Wozel 2001). Calcipotriol monotherapy was also less effective than treatment with calcipotriol to which halometasone was added (SMD 1.13; 95% CI 0.64 to 1.62) (Yang 2009). Monotherapy with tacalcitol (8 weeks) was less effective than sequential treatment with a combined calcipotriol and betamethasone dipropionate product (4 weeks) followed by calcipotriol monotherapy for a further 4 weeks (SMD 0.49; 95% CI 0.31 to 0.67) (Ortonne 2004). White 2006 (P) compared three regimens, all of which included an initial phase in which participants applied combination treatment with calcipotriol and betamethasone dipropionate once a day for four weeks. The subsequent eight‐week maintenance phase using placebo ointment was significantly less effective than maintenance with either twice‐daily calcipotriol (SMD 0.25; 95% CI 0.10 to 0.39), or maintenance with calcipotriol on weekdays and combination treatment at weekends (SMD 0.59; 95% CI 0.45 to 0.74). When the two active maintenance regimens were compared directly, the alternating (weekday/weekend) regimen was significantly more efficacious (SMD 0.30; 95% CI 0.16 to 0.45) (White 2006 (H)).

In the remaining four intervention‐comparator contrasts, we did not detect any significant difference in the PASI assessments (see Analysis 13.3). PASI assessments in the study by Kragballe 2004 found the difference between two complex regimens not to be significant (SMD 0.15; 95% CI ‐0.01 to 0.30). This result was on the borderline of significance and contrasted with the IAGI assessment in the same study, which was statistically significant in favour of the complex regimen (see Analysis 13.1).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

We did not identify any relevant study that provided PASI data for this analysis.

Analysis 15: Vitamin D analogues versus other treatment

This comparison incorporated all other vitamin D head‐to‐head comparisons of treatments for psoriasis of the body (excluding inverse psoriasis) that had not already been included (see Analysis 15.3 and Table 18). We included 12 intervention‐comparator contrasts, with PASI data available for 6 of these contrasts. There were six between‐patient trials and three within‐patient trials, with data for 1228 participants. Trial duration ranged between 4 and 12 weeks. In light of the pharmacological diversity of the comparators, we only pooled data within subgroups (Higgins 2011).

According to the PASI assessment, twice‐daily calcipotriol was significantly more effective than coal tar polytherapy (SMD ‐0.63; 95% CI ‐1.06 to ‐0.20; I² statistic = NA) as well as propylthiouracil cream (SMD ‐2.24; 95% CI ‐3.23 to ‐1.25; I² statistic = NA). Twice‐daily calcipotriol was not significantly more effective than coal tar (SMD ‐0.10; 95% CI ‐1.54 to 1.35; I² statistic = 92.8%), betamethasone dipropionate and salicylic acid (SMD ‐0.05; 95% CI ‐0.36 to 0.26; I² statistic = NA), or vitamin B12 cream (SMD ‐0.01; 95% CI ‐0.78 to 0.75; I² statistic = NA). The high level of heterogeneity observed in the pooled analysis of calcipotriol versus coal tar reflects contradictory findings from two studies, one finding in favour of calcipotriol (Tham 1994) and the other in favour of coal tar (Alora‐Palli 2010).

When compared with once‐daily dosing, twice‐daily vitamin D was borderline in demonstrating a significantly better effect (SMD ‐0.12; 95% CI ‐0.25 to 0.00; I² statistic = 0%). The subgroups contributing to this result were dosing comparisons of calcipotriol (SMD ‐0.12; 95% CI ‐0.28 to 0.03; I² statistic = 0%) and combination treatment with calcipotriol and betamethasone dipropionate (SMD ‐0.12; 95% CI ‐0.32 to 0.09; I² statistic = NA).

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of topical treatments for inverse or facial psoriasis (see Analysis 16.3 and Table 19). We found evidence on four treatments in this comparison: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus. We only identified one placebo‐controlled trial evaluating tacrolimus ointment (Lebwohl 2004), but the study did not report any effectiveness data suitable for this review. However, the study did contribute data on adverse events and withdrawal rates.

PASI data were available for 3 of the 4 topical treatments for inverse psoriasis, all reported by 1 4‐week between‐patient study with 75 participants (Kreuter 2006 (P)). In this study, participants applied all treatments once daily. The SMD for the PASI found a statistically significant difference in favour of betamethasone valerate 0.1% (SMD ‐2.83; 95% CI ‐3.79 to ‐1.88) as well as calcipotriol ointment (SMD ‐1.08; 95% CI ‐1.77 to ‐0.40). However, PASI data on once‐daily pimecrolimus cream indicated the difference relative to vehicle was not statistically significant (SMD ‐0.62; 95% CI ‐1.27 to 0.02).

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for inverse psoriasis, where vitamin D was compared with an active control (see Analysis 17.3 and Table 20). We identified five intervention‐comparator contrasts. We compared four treatments with calcipotriol: once‐daily betamethasone valerate, combined treatment with calcipotriol and hydrocortisone, calcitriol, and pimecrolimus. Calcitriol was compared with tacrolimus.

Two between‐patient studies contributed PASI data from 464 study participants on 3 of the 5 intervention‐comparator contrasts. Kreuter 2006 (H) found that calcipotriol was significantly less effective than betamethasone valerate (SMD 2.02; 95% CI 1.20 to 2.84), but not significantly different in effect to pimecrolimus (SMD ‐0.53; 95% CI ‐1.17 to 0.11). Participants in the Kreuter 2006 (H) trial applied all treatments once daily. When assessed using the PASI, Ortonne 2010 found that calcipotriol was significantly less effective on inverse psoriasis than combined treatment with calcipotriol and hydrocortisone (SMD 0.32; 95% CI 0.12 to 0.51). This finding was consistent with the IAGI assessment in the same trial (see Analysis 17.1).

Analysis 18: Scalp psoriasis: placebo‐controlled trials

No PASI data were available for this comparison.

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

No PASI data were available for this comparison.

Analysis 1: Vitamin D analogues versus placebo

Our review included eight vitamin D analogues for body psoriasis in this comparison, with PAGI data available for three treatments (see Analysis 1.4 and Table 5). Five studies reported PAGI data, all of which were between‐patient design, and 1467 participants contributed data. In all six studies, treatment duration was eight weeks. The pooled effect across the 3 studies reporting PAGI data was SMD ‐0.54; 95% CI ‐0.72 to ‐0.36; I² statistic = 55.5%.

Analysis 2: Corticosteroid (potent) versus placebo

Our review did not identify any study that compared potent corticosteroids against placebo and reported PAGI data.

Analysis 3: Corticosteroid (very potent) versus placebo

We included three very potent corticosteroids for body psoriasis in this comparison for this outcome (see Analysis 3.4 and Table 7). Three studies reported PAGI data for 283 participants on 2 of these 3 treatments. There was one between‐patient trial (Lebwohl 2002) and two within‐patient studies. In all three studies, treatment duration was two weeks. The pooled effect across both treatments for PAGI was (SMD ‐1.22; 95% CI ‐1.42 to ‐1.02; I² statistic = 0%). Both treatments performed statistically significantly better than placebo. PAGI data on the use of very potent steroids on the scalp are reported in Analysis 18.4.

Analysis 4: Dithranol versus placebo

Our review did not identify any study comparing dithranol against placebo and that reported PAGI data.

Analysis 5: Vitamin D combination products versus placebo

This comparison included treatment with combined calcipotriol and betamethasone dipropionate used either once or twice daily on the body (see Analysis 5.4 and Table 8). None of the studies of twice‐daily treatment reported PAGI data, but one parallel‐group study with 235 participants contributed data on once‐daily treatment (Langley 2011 (P)). Treatment duration was eight weeks. The PAGI SMD was ‐0.69 (95% CI ‐0.98 to ‐0.40; I² statistic = NA). Placebo‐controlled trials of combination vitamin D/steroid treatments for scalp psoriasis are reported in Analysis 18.4.

Analysis 6: Other treatment versus placebo

This comparison comprised all other treatments for body psoriasis that we did not include in the first five comparisons; therefore, we removed pooling. None of the studies assessed the same treatment, which means that findings should be interpreted with caution.

In total, we included 26 treatments in this analysis (see Analysis 6.4 and Table 9). Two studies with 105 participants reported PAGI data on 2 of these 26 treatments. Both trials were between‐patient studies. Treatment duration ranged from 3 to 12 weeks. According to participants' assessments (PAGI), betamethasone 17‐valerate 21 acetate plus tretinoin plus salicylic acid performed significantly better than placebo (SMD ‐0.80; 95% CI ‐1.26 to ‐0.35), whilst kukui nut oil was no more effective than placebo (SMD 0.00; 95% CI ‐0.80 to 0.80). These findings were very similar to the investigators' assessments (see Analysis 6.1).

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

Our review identified eight vitamin D analogue‐potent corticosteroid comparisons for body psoriasis (see Analysis 7.4 and Table 10). Two studies with 738 participants reported PAGI data for 1 of these 8 comparisons. One trial was a parallel‐group study (Cunliffe 1992), and one was a within‐patient study (Kragballe 1991a), both with a treatment duration of six weeks. The PAGI SMD indicated that the vitamin D analogue calcipotriol was significantly more effective than betamethasone valerate (SMD ‐0.26; 95% CI ‐0.38 to ‐0.14; I² statistic = 0%).

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

This comparison included one vitamin D analogue and a very potent corticosteroid contrast, calcipotriol ointment versus clobetasol propionate foam for body psoriasis (see Table 11 and Analysis 8.4). One study with 42 participants reported PAGI data. Koo 2006 was a between‐patient study with a treatment duration of two weeks. Findings for the participant‐reported assessment (SMD 0.42; 95% CI ‐0.20 to 1.03) were similar, the investigators' assessment from the same study (SMD 0.19; 95% CI ‐0.42 to 0.80; see Analysis 8.1).

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

We found one study comparing a vitamin D analogue‐steroid combination against corticosteroids for body psoriasis that reported PAGI/PGA data (see Table 12 and Analysis 9.4). This two‐week between‐patient study reported PAGI data for 65 participants (Koo 2006), comparing combined treatment with calcipotriol and clobetasol propionate against the very potent corticosteroid alone. According to the participants' assessment (PAGI), there was no statistically significant difference between the treatment options (SMD ‐0.28; 95% CI ‐0.80 to 0.24). This finding contrasts with the investigators' assessment (IAGI), which found in favour of combination treatment (SMD ‐0.69; 95% CI ‐1.22 to ‐0.15) (see Analysis 9.1).

Analysis 10: Vitamin D alone or in combination versus dithranol

This comparison considered vitamin D analogues against dithranol (see Analysis 10.4 and Table 13). We identified three intervention‐comparator contrasts: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. PAGI data were available only for the intervention‐comparator contrast of calcipotriol versus dithranol. Two between‐patient trials reported PAGI data for 544 participants (Berth Jones 1992b; Van de Kerkhof 2006). Treatment duration ranged from 8 to 12 weeks. The SMD for the PAGI was ‐0.05 (95% CI ‐0.90 to 0.80; I² statistic = 92.5%), indicating no statistically significant advantage for calcipotriol or dithranol according to the participants' assessment.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Our review identified three intervention‐comparator contrasts for body psoriasis in this comparison: calcipotriol versus calcitriol, calcipotriol versus tacalcitol, and calcipotriol versus maxacalcitol (see Analysis 11.4 and Table 14). One between‐patient trial involving 250 participants contributed PAGI data for the comparison of calcipotriol and calcitriol (Ji 2008). Treatment duration was 12 weeks. The SMD for the PAGI indicated that there was no statistically significant difference between calcipotriol and calcitriol (SMD 0.04; 95% CI ‐0.21 to 0.29; I² statistic = NA). This study used three measures to assess the effects of calcipotriol and calcitriol. The investigators' assessment supported the participant assessment (no difference) (see Analysis 11.1), but the TSS found a significant difference in favour of calcipotriol (Analysis 11.2).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

Our review identified 12 intervention‐comparator contrasts for body psoriasis, involving 3 vitamin D analogues, 2 combination products, and 7 different corticosteroids (see Analysis 12.4 and Table 15). Two parallel‐group trials contributed PAGI data from 399 participants for 2 of these 12 intervention‐comparator contrasts. Treatment duration ranged between two and eight weeks. According to the participant assessment, twice‐daily calcipotriol was significantly less effective than twice‐daily treatment with calcipotriol and clobetasol propionate (SMD 0.70; 95% CI 0.16 to 1.23). Treatment with tacalcitol was also less effective than treatment with a combined product containing calcipotriol and betamethasone dipropionate, when both treatments were applied once daily (SMD 0.46; 95% CI 0.24 to 0.68). The pooled SMD for these two trials (0.49; 95% CI 0.29 to 0.69; I² statistic = 0%) gave greater weight to the comparison with tacalcitol, since this was the larger trial (334 participants; Langley 2011 (H)).

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison summarises findings on complex regimens for body psoriasis, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments (see Analysis 13.4 and Table 16). We identified 12 intervention‐comparator contrasts and data from the Patient Assessment of Global Improvement (PAGI) or Patient Global Assessment of Disease Severity (PGA) were available for 7 of these. Data from 2508 participants contributed to the PAGI analysis, based on findings from 4 between‐patient trials. Trial duration varied between 8 and 12 weeks. Because the interventions and comparators were highly variable and two trials each contributed three pair‐wise contrasts, we did not pool the data.

Four intervention‐comparator contrasts for which PAGI data were available found a significant difference between regimens. Ortonne 2004 found that four weeks of treatment with a combination product of calcipotriol and betamethasone dipropionate, followed by monotherapy with calcipotriol for four weeks, was significantly more effective than monotherapy with tacalcitol (SMD 0.54; 95% CI 0.36 to 0.72). Findings from the patient‐reported outcome were identical to the investigators' assessment (Analysis 13.1). The trial by White 2006 (H)/White 2006 (P) compared three regimens. All included an initial phase in which participants applied once‐daily combination treatment with calcipotriol and betamethasone dipropionate for four weeks, but differed in their subsequent eight‐week maintenance phase. According to the participants' assessment, maintenance therapy with twice‐daily placebo ointment was significantly less effective than maintenance with either calcipotriol ointment (SMD 0.28; 95% CI 0.13 to 0.42) or maintenance with calcipotriol on weekdays and combination treatment at weekends (SMD 0.71; 95% CI 0.56 to 0.85). We compared the 2 regimens with active maintenance options directly; the alternating (weekday/weekend) approach was significantly more effective (SMD 0.44; 95% CI 0.29 to 0.58). These findings, based on the participants' assessment of severity, were aligned with those of both the investigators' assessment (Analysis 13.1) and the PASI assessment (Analysis 13.3).

A head‐to‐head comparison of two complex regimens found a significant difference (Kragballe 2004). Once‐daily combination treatment (calcipotriol/betamethasone dipropionate) for 8 weeks followed by once‐daily calcipotriol for 4 weeks was significantly less effective than a regimen consisting of combined treatment for 4 weeks, followed by 8 weeks of once‐daily therapy with calcipotriol on weekdays and combined therapy at weekends (SMD 0.23; 95% CI 0.07 to 0.39).

In three of the seven intervention‐comparator contrasts, no significant difference was found. Kragballe 2004 also compared 12 weeks of calcipotriol monotherapy with the 2 complex regimens described in the paragraph above. Compared with monotherapy, participants who applied once‐daily combination treatment (calcipotriol/betamethasone dipropionate) for 8 weeks followed by once‐daily calcipotriol for 4 weeks did not experience a significantly greater improvement in their psoriasis (SMD ‐0.14; 95% CI ‐0.30 to 0.02). Similarly, the psoriasis of participants on monotherapy with twice‐daily calcipotriol did not improve significantly differently compared to those who applied once‐daily combination treatment (calcipotriol/betamethasone dipropionate) for 4 weeks, followed by 8 weeks using calcipotriol once daily (weekdays) and combination therapy (weekends) (SMD 0.10; 95% CI ‐0.06 to 0.26).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

We did not identify any relevant study that provided PAGI data for this analysis.

Analysis 15: Vitamin D analogues versus other treatment

This comparison incorporated all other vitamin D head‐to‐head comparisons of treatments for psoriasis of the body (excluding inverse psoriasis) that we had not already included (see Analysis 15.4 and Table 18). We included 12 intervention‐comparator contrasts, with PAGI data available for 6 of these contrasts: 3 between‐patient trials and 3 within‐patient trials data for 456 participants. Trial duration ranged between 4 and 12 weeks. In light of the pharmacological diversity of the comparators, we only pooled data within subgroups.

According to the participants' assessment, twice‐daily calcipotriol was significantly more effective than three comparators, with evidence on each based on a single trial. Calcipotriol was more effective than coal tar both as monotherapy (SMD ‐1.51; 95% CI ‐2.12 to ‐0.90; Tham 1994) and as polytherapy (SMD ‐0.56; 95% CI ‐0.99 to ‐0.13; Van de Kerkhof 2002a), and more effective than betamethasone dipropionate and salicylic acid (SMD ‐0.49; 95% CI ‐0.79 to ‐0.20; Scarpa 1994). No significant difference was evident when calcipotriol was compared to tacrolimus ointment (Ortonne 2006), tazarotene (Tzung 2005), or vitamin B12 cream (Stuecker 2001).

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of topical treatments for inverse or facial psoriasis (see Analysis 16.4 and Table 19). We found evidence on four treatments in this comparison: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus. We identified only one placebo‐controlled trial evaluating tacrolimus ointment (Lebwohl 2004), but the study did not report any effectiveness data suitable for this review. However, the study did contribute data on adverse events and withdrawal rates.

PAGI (patient‐assessment) data were available for one of the four topical treatments for inverse psoriasis. One 8‐week between‐patient study (Gribetz 2004) reported data from 47 participants. Relative to placebo, the SMD for the PAGI found a statistically significant difference in favour of twice‐daily pimecrolimus cream (SMD ‐0.65; 95% CI ‐1.24 to ‐0.06).

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

We did not identify any relevant study that provided PAGI data for this analysis.

Analysis 18: Scalp psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of treatments for scalp psoriasis (see Analysis 18.4 and Table 21). We included evidence on 11 treatments in this comparison, with PAGI data available for 5 treatments. Five between‐patient trials contributed data from 1875 participants. Trial duration ranged between three and eight weeks.

Four of the five treatments for scalp psoriasis that were evaluated using the Patient Global Assessment Scale were significantly more effective than placebo. The least effective treatment was calcipotriol (SMD ‐0.66; 95% CI ‐1.28 to ‐0.05; I² statistic = 74.5%), and the most effective was betamethasone dipropionate (SMD ‐1.23; 95% CI ‐1.43 to ‐1.03; I² statistic = NA). Effects for the very potent steroid amcinonide (SMD ‐0.97; 95% CI ‐1.33 to ‐0.61 I² statistic = NA) and combination treatment with calcipotriol and betamethasone dipropionate (SMD ‐1.00; 95% CI ‐1.79 to ‐0.22; I² statistic = 93.2%) fell between these two extremes. There were no data suitable for pooling across subcategories.

Only for ciclopirox olamine shampoo was the difference relative to placebo found to be non‐significant: SMD ‐0.11 (95% CI ‐0.86 to 0.64; I² statistic = NA).

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.4 and Table 22).

We identified six intervention‐comparator contrasts, and PAGI data were available for four of these contrasts (there were no participant‐assessed outcomes data for the comparison of calcipotriol against either clobetasol propionate or against coal tar polytherapy). All studies were parallel‐group in design (between‐patient). Six studies contributed PAGI data from 3742 participants, and trial duration ranged from 4 to 8 weeks.

Based on the participant assessment scores, calcipotriol was significantly less effective than betamethasone dipropionate (SMD 0.56; 95% CI 0.31 to 0.81; I² statistic = 82.8%), betamethasone valerate (SMD 0.41; 95% CI 0.22 to 0.59; I² statistic = NA), and combination treatment with calcipotriol and betamethasone dipropionate (SMD 0.84; 95% CI 0.61 to 1.08; I² statistic = 81.5%). Combination treatment (calcipotriol/betamethasone dipropionate) was significantly more effective than betamethasone dipropionate alone (SMD ‐0.17; 95% CI ‐0.25 to ‐0.09; I² statistic = 0%).

Analysis 1: Vitamin D analogues versus placebo

Our review included eight vitamin D analogues for body psoriasis in this comparison (see Analysis 1.5 and Table 5). Thirty studies, involving 4986 participants, contributed data. Eighteen trials were between‐patient design, and 12 were within‐patient studies. Treatment duration ranged from 3 to 12 weeks. Six studies had adequately concealed treatment allocation. The pooled effect across all 30 studies was ‐0.90 (95% CI ‐1.07 to ‐0.72; I² statistic = 87.5%), which equates to 1.03 on a 6‐point IAGI scale. However, the summary statistic conceals considerable variation between treatments. In two treatments, the effect was not statistically significantly different to placebo (twice‐daily calcipotriol plus occlusion for 2 weeks followed by 4 weeks with no treatment, and once‐daily becocalcidiol). In treatments that were significantly more effective than placebo, the effect ranged from ‐0.67 (twice‐daily becocalcidiol) to ‐1.66 (once‐daily paricalcitol); on a 6‐point IAGI scale, these effects translate into 0.80 and 1.91 points, respectively. However, given the high level of heterogeneity found in the meta‐analysis, it may be more informative to look at individual products within the class, rather than treating this class as a single group. We removed the pooling across the class, so that we only pooled subtotals.

We reported placebo‐controlled trials of vitamin D products for scalp psoriasis (Analysis 18.5) and inverse psoriasis (Analysis 16.5) elsewhere.

Analysis 2: Corticosteroid (potent) versus placebo

Our review identified 10 potent corticosteroids for body psoriasis in this comparison (see Analysis 2.5 and Table 6). The treatments included three betamethasone dipropionate regimens, but we identified no effectiveness evidence for budesonide and no study reporting PAGI data. Therefore, 13 studies involving 2216 participants contributed data on 9 of the 10 treatments. Eleven trials were between‐patient design, and two were within‐patient studies. Treatment allocation was adequately concealed in four studies. Treatment duration ranged from 2 to 12 weeks. Participant numbers for individual studies ranged from 9 (Wortzel 1975 (2)) to 633 (Kaufmann 2002 (P)). The pooled effect across all 13 studies was ‐0.89 (95% CI ‐1.06 to ‐0.72; I² statistic = 65.1%). This equates to 1.02 on a 6‐point IAGI scale. With the exception of diflorasone diacetate (Lane 1983), all potent corticosteroids were significantly more effective than placebo at the 5% level of significance. We reported elsewhere placebo‐controlled trials of potent corticosteroid products for scalp psoriasis (Analysis 18.5) and inverse psoriasis (Analysis 16.5).

Analysis 3: Corticosteroid (very potent) versus placebo

Our review identified three very potent corticosteroids for body psoriasis in this comparison (see Analysis 3.5 and Table 7), but effectiveness data were not available for one of these treatments (halcinonide). Ten studies reported data for 1264 participants. There were seven between‐patient trials and three within‐patient studies. Treatment duration ranged from two to four weeks. In no study was treatment allocation adequately concealed. The SMD across the 2 treatments was ‐1.56 (95% CI ‐1.87 to ‐1.26; I² statistic = 81.7%; 10 studies). This equates to 1.80 points on a 6‐point IAGI scale. Both clobetasol propionate (SMD ‐1.65; 95% CI ‐2.10 to ‐1.20; I² statistic = 86.3%; 7 studies) and halobetasol (SMD ‐1.36; 95% CI ‐1.65 to ‐1.07 I² statistic = 47.1%; 3 studies) performed significantly better than placebo.

We reported elsewhere placebo‐controlled trials of very potent corticosteroid products for scalp psoriasis (Analysis 18.5).

Analysis 4: Dithranol versus placebo

This comparison considered dithranol against placebo (see Analysis 4.5 and Table 23). Three within‐patient trials reported data for 47 participants. The only outcome reported by these trials was the TSS, so results for the combined end point are identical to the TSS results section. The types of treatment comprised the following: dithranol 0.1% in a carbamide (17% urea) base twice daily (Buckley 1978), dithranol in aqueous gel (dose titration 0.1% to 2.0%) twice daily (Grattan 1997 (P)), and dithranol 2% ointment 'one minute therapy' once daily (Jekler 1992).

Treatment duration ranged from three to eight weeks. In all three studies, the adequacy of the concealment of treatment allocation was unclear. The SMD for the combined end point was ‐1.06 (95% CI ‐1.66 to ‐0.46; I² statistic = 37.4%; 3 studies), which equates to 1.22 on a 6‐point IAGI scale. All three trials found a statistically significant effect in favour of dithranol.

Analysis 5: Vitamin D combination products versus placebo

This comparison included treatment with combined calcipotriol and betamethasone dipropionate used either once or twice daily on the body (see Analysis 5.5 and Table 8). As all five studies reported IAGI data, results for the combined end point are identical to those in Analysis 5.1. Five parallel‐group studies with 2058 participants contributed data. Treatment duration ranged from four to eight weeks. Three trials adequately concealed treatment allocation. The SMD for the combined end point across treatments was ‐1.44 (95% CI ‐1.76 to ‐1.12; I² statistic = 89.4%; 5 studies), with twice‐daily combination treatment (SMD ‐1.90; 95% CI ‐2.09 to ‐1.71; 2 studies) achieving a significantly larger effect than once‐daily treatment (SMD ‐1.21; 95% CI ‐1.50 to ‐0.91; 3 studies). On a 6‐point IAGI scale, these equate to improvements of 2.18 and 1.39 points, respectively.

However, the PASI assessments differed from the IAGI: although twice‐daily combination treatment achieved a larger effect than once‐daily treatment (SMD ‐1.41 and ‐1.14, respectively), the difference was not statistically significant (see Analysis 5.3).

In Analysis 18.5, we reported placebo‐controlled trials of combination vitamin D/steroid treatments for scalp psoriasis.

Analysis 6: Other treatment versus placebo

This comparison comprised all other treatments for psoriasis of the body that we did not include in the first five comparisons; therefore, we removed pooling. None of the studies assessed the same treatment, which means that findings should be interpreted with caution.

In total, we included 26 treatments in this analysis (see Analysis 6.5 and Table 9). Twenty‐six studies, with 1450 participants, reported data on 25 of these 26 treatments. The study of omega‐3‐polyunsaturated fatty acids ointment (Henneicke‐v. Z. 1993) did not report usable effectiveness data, but did contribute data to the analysis of withdrawals. Twelve trials were between‐patient design, and 14 were within‐patient studies. Treatment duration ranged from 2 to 12 weeks. Two studies adequately concealed treatment allocation (Levine 2010 (P); Stutz 1996).

As assessed by the combined end point, just over half the treatments (13/25) performed statistically significantly better than placebo: aloe vera cream, anti‐IL‐8 monoclonal antibody cream, betamethasone 17‐valerate 21‐acetate plus tretinoin plus salicylic acid, combined treatment with calcipotriene and nicotinamide, fish oil plus occlusion, herbal skin care products, indigo naturalise 1.4% ointment, Mahonia aquifolium, methotrexate gel, mycophenolic acid ointment, PTH (1‐34) in Novasome cream®, tazarotene, and theophylline 1% ointment. The effect size (SMD) for the combined end point ranged from ‐0.48 (Levine 2010 (P); calcipotriene combined with nicotinamide) to ‐2.96 (Maier 2004; herbal skin care products), or 0.55 to 3.40, respectively, on a 6‐point IAGI scale.

The trial of herbal skin care products enrolled just 34 participants, of which 24 contributed data (Maier 2004), so this study is classed as having a high risk of attrition bias (Figure 3; see Risk of bias in included studies). The results of this study were published as an abstract, and our searches failed to locate a full publication. Therefore, findings should be treated with caution.

In the remaining 12 treatments, the difference relative to placebo using the combined end point was not statistically significant. These included topical caffeine, Dead Sea salts emollient lotion, hexafluoro‐1,25‐dihydroxyvitamin D3, kukui nut oil, NG‐monomethyl‐L‐arginine (L‐NMMA) cream, nicotinamide 1.4%, oleum horwathiensis, platelet aggregation activating factor (PAF), polymyxin B cream, topical sirolimus, topical tacrolimus, and tar.

In two treatments, findings by different outcomes were inconsistent. When assessed by the TSS, hexafluoro‐1,25‐dihydroxyvitamin D3 was significantly more effective than placebo (‐1.13; 95% CI ‐1.91 to ‐0.35) (Analysis 6.2), whereas the difference assessed by the IAGI was non‐significant (‐0.62; 95% CI ‐1.35 to 0.12) (Analysis 6.1). We based these findings on a single trial (Durakovic 2001). Similarly, the difference for oleum horwathiensis was statistically significant using the TSS (‐0.77; 95% CI ‐1.40 to ‐0.14) (Analysis 6.2) whereas the difference assessed by the IAGI was not (‐0.02; 95% CI ‐0.63 to 0.58) (Analysis 6.1). We also based findings for this treatment on a single trial (Lassus 1991).

We reported elsewhere placebo‐controlled trials of other treatments for scalp psoriasis (Analysis 18.5) and inverse psoriasis (Analysis 16.5).

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

Our review identified eight vitamin D analogue‐potent corticosteroid comparisons for body psoriasis for this comparison (see Analysis 7.5 and Table 10). Fourteen studies with 3542 participants reported data for these 8 intervention‐comparator contrasts. Nine trials were between‐patient design, and five were within‐patient studies. Treatment duration ranged from three to eight weeks. One study adequately concealed treatment allocation (Papp 2003 (H)).

The combined end point SMD indicated that, overall, there was no statistically significant difference between vitamin D and potent corticosteroids: SMD 0.11 (95% CI ‐0.07 to 0.30; I² statistic = 85.6%; 14 studies). There was however substantial heterogeneity and variation in effect underlying this summary statistic. Therefore, we only pooled subtotals.

The difference between the vitamin D analogue and the potent corticosteroid was not statistically significant for four intervention‐comparator contrasts (calcipotriol versus betamethasone valerate, calcipotriol versus desoxymetasone, calcitriol versus betamethasone dipropionate, calcitriol versus betamethasone valerate). In the comparison of calcipotriol and betamethasone valerate, the non‐significant result was borderline (SMD ‐0.12; 95% CI ‐0.26 to 0.02; I² statistic = 41.6%; 4 studies). When assessed using the TSS (one study), PASI (four studies), and PAGI (two studies), calcipotriol was significantly more effective than betamethasone valerate (see Analysis 7.2, Analysis 7.3, and Analysis 7.4). There were similar inconsistencies between the outcome assessments for the comparison of calcitriol with betamethasone dipropionate, all of which we based on a single study (Camarasa 2003). According to the TSS and PASI, calcitriol was significantly less effective than betamethasone dipropionate, but the IAGI indicated that the difference was not significant (see Analysis 7.1).

In one intervention‐comparator contrast (calcipotriol versus fluocinonide), the vitamin D analogue was significantly more effective than potent corticosteroid: SMD ‐0.58 (95% CI: ‐0.99 to ‐0.18; I² statistic = NA). This effect is equivalent to a gain of approximately two‐thirds of a point (0.64) on a 6‐point IAGI scale. In three intervention‐comparator contrasts, the potent corticosteroid was significantly more effective than the vitamin D analogue (calcipotriol versus betamethasone dipropionate, calcipotriol versus diflorasone diacetate, tacalcitol versus betamethasone valerate). The SMD effect sizes for these were 0.43, 0.27, and 0.41, respectively, equivalent to improvements of 0.47, 0.30, and 0.45 on a 6‐point IAGI scale.

We reported elsewhere trials comparing vitamin D and potent corticosteroids for scalp psoriasis (Analysis 19.5) and inverse psoriasis (Analysis 17.5).

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

This comparison considered vitamin D analogues against very potent corticosteroids for body psoriasis (see Analysis 8.5 and Table 11). We found data on one intervention‐comparator contrast: calcipotriol versus clobetasol propionate. Two between‐patient trials reported data for 82 participants. Treatment duration ranged between two and six weeks. The adequacy of treatment allocation was unclear in both studies. The SMD for the combined end point indicated that there was no statistically significant difference between the vitamin D analogue and the very potent corticosteroid: ‐0.06 (95% CI ‐0.57 to 0.44; I² statistic = 25.7%; 2 studies).

We reported elsewhere trials comparing vitamin D and very potent corticosteroids for scalp psoriasis (Analysis 19.5).

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

This comparison considered combined treatment with vitamin D analogues and corticosteroids against potent or very potent corticosteroid for body psoriasis (see Analysis 9.5 and Table 12). Five parallel‐group studies, ranging in treatment duration from 2 to 8 weeks, provided data on 2113 participants. The adequacy of the concealment of treatment allocation was unclear in all five trials.

Effectiveness data were available for three contrasts:

• calcipotriol plus betamethasone dipropionate versus betamethasone dipropionate;

• calcipotriol plus betamethasone dipropionate versus clobetasol propionate; and

• calcipotriol plus clobetasol propionate versus clobetasol propionate.

Overall, combination treatment was more effective than monotherapy with corticosteroids: SMD ‐0.26 (95% CI ‐0.52 to ‐0.00; I² statistic = 84.4%; 5 studies). This effect is equivalent to a change of 0.29 on a 6‐point IAGI scale. However, the pooled finding masks important diversity between the three intervention‐comparator contrasts. When we compared combination treatment with calcipotriol and betamethasone dipropionate with a potent steroid (betamethasone dipropionate alone), combination treatment was more effective (SMD ‐0.40; 95% CI ‐0.52 to ‐0.27; I² statistic = 41.8%; 3 studies). However, the same combination treatment was significantly less effective than monotherapy with a very potent steroid (clobetasol propionate) (SMD 0.45; 95% CI 0.09 to 0.81; I² statistic = NA). Calcipotriol in combination with clobetasol propionate however was more effective than the very potent steroid used alone (SMD ‐0.69; 95% CI ‐1.22 to ‐0.15; I² statistic = NA). Therefore, we removed pooling for this analysis, but the findings demonstrate that combination treatment with vitamin D analogue and a corticosteroid is more effective than the same steroid used as a monotherapy.

We reported elsewhere trials comparing combination therapy (with vitamin D and potent corticosteroids) against potent steroids for scalp psoriasis (Analysis 19.5).

Analysis 10: Vitamin D alone or in combination versus dithranol

This comparison considered vitamin D analogues against dithranol for body psoriasis (see Analysis 10.5 and Table 13). We identified three intervention‐comparator contrasts: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. Seven between‐patient trials and one within‐patient trial (Grattan 1997 (H)) reported data for 1284 participants. We considered concealment of treatment allocation to be adequate in one trial (Van de Kerkhof 2006). Treatment duration ranged from 4 weeks to 12 weeks. In addition, there was some variation in the dithranol regimens employed by trials and in the baseline severity of trial participants. These factors may explain the high level of heterogeneity found in the pooled results.

The SMD for the combined end point was 0.09 (95% CI ‐0.44 to 0.63; I² statistic = 94.9%; 8 studies), indicating that there was no evidence of a statistically significant difference in effect between vitamin D analogues and dithranol. The pooled findings for calcipotriol against dithranol (SMD 0.07; 95% CI ‐0.57 to 0.71; I² statistic = 95.7%; 6 studies) and tacalcitol versus dithranol (SMD ‐0.18; 95% CI ‐0.60 to 0.25; I² statistic = NA) were aligned with, and drove, this result. However, the high levels of heterogeneity mean that findings merit closer scrutiny. Six studies contributed data from 1086 participants to the comparison of calcipotriol and dithranol. In three of these studies, calcipotriol performed significantly better than dithranol; two trials found significant differences in favour of dithranol; and one study found no significant difference between the treatments (see Analysis 10.5). A single study of 114 participants compared calcitriol against dithranol (Hutchinson 2000) and found a statistically significant difference in favour of dithranol (SMD 0.51; 95% CI 0.14 to 0.88; I² statistic = NA). This equates to just over half a point (0.56 of a point) improvement on a 6‐point IAGI scale. However, the same study found no significant difference between the treatments in the assessment using the TSS (Analysis 10.2) or the PASI (Analysis 10.3).

In light of the considerable level of observed variation within and between studies (Higgins 2011), we removed pooling from this comparison.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Our review identified three intervention‐comparator contrasts for body psoriasis in this comparison: calcipotriol versus calcitriol, calcipotriol versus tacalcitol, and calcipotriol versus maxacalcitol (see Analysis 11.5 and Table 14). Four trials involving 513 participants contributed data. One trial was within‐patient (Barker 1999 (H)); three were between‐patient in design; and no trials demonstrated the concealment of treatment allocation to be adequate. Treatment duration ranged from 8 to 12 weeks.

The SMD for the combined end point indicated that there was no statistically significant difference between the treatments: ‐0.17 (95% CI ‐0.62 to 0.27; I² statistic = 78.5%; 4 studies). When we considered individual intervention‐comparator contrasts using the combined end point, calcipotriol was more effective than tacalcitol (SMD ‐0.47; 95% CI ‐0.73 to ‐0.21; I² statistic = NA), which equates to an improvement of 0.52 on a 6‐point IAGI scale. However, there was no statistically significant difference between calcipotriol and calcitriol (SMD ‐0.41; 95% CI ‐1.46 to 0.64; I² statistic = 72.3%; 2 studies) or between calcipotriol and maxacalcitol (SMD 0.43; 95% CI ‐0.12 to 0.98; I² statistic = NA). Participants in all trials applied treatments twice daily, with the exception of tacalcitol, which was applied once daily (Veien 1997).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

Our review identified 12 intervention‐comparator contrasts for body psoriasis, involving 3 vitamin D analogues, 2 combination products, and 7 different corticosteroids (see Analysis 12.5 and Table 15). Seventeen trials involving 5856 participants contributed data. Sixteen trials were between‐patient, and 1 was within‐patient in design (Salmhofer 2000). Treatment duration ranged from 2 to 12 weeks. Four trials adequately concealed treatment allocation, but concealment was inadequate in one trial (in the other trials, insufficient data were reported to allow an assessment of adequacy).

Overall, vitamin D plus corticosteroid was more effective than vitamin D alone: The SMD for the combined end point was 0.46 (95% CI 0.33 to 0.59; I² statistic = 83.3%; 17 studies), which translates into half of 1 point (0.50) on a 6‐point IAGI scale. The substantial level of heterogeneity reflects not only different dosing schedules of the intervention and comparator products, but also the different potency of the corticosteroids and the different vitamin D analogues evaluated; we therefore removed pooling (Higgins 2011).

Twice‐daily calcipotriol was significantly less effective than combination treatment with betamethasone dipropionate. This finding held whether participants applied the corticosteroid separately at night time (SMD 0.56; equivalent to 0.61 on a 6‐point IAGI scale) or as a combined product used once daily or twice daily (SMD 0.43 and 0.66, respectively, translating into improvements of 0.48 and 0.73 points on a 6‐point IAGI scale). Twice‐daily calcipotriol was also significantly less effective than combination treatment with clobetasone butyrate (SMD 0.27; 95% CI 0.05 to 0.48; I² statistic = NA; 0.29 of a point on a 6‐point IAGI) or clobetasol propionate (SMD 0.88; 95% CI 0.34 to 1.42; I² statistic = NA; 0.97 of a point on a 6‐point IAGI scale). Once‐daily combination treatment with calcipotriol and betamethasone dipropionate was significantly more effective than once‐daily treatment with either calcipotriol (SMD 0.66; 95% CI 0.31 to 1.02; I² statistic = 80.9%; 2 studies) or tacalcitol (SMD 0.48; 95% CI 0.26 to 0.70; I² statistic = NA).

In none of the 12 intervention‐comparator contrasts was the vitamin D analogue significantly more effective than combined vitamin D plus corticosteroid. However, in five instances, there was no significant difference between vitamin D and the comparator (combination treatment). These included twice‐daily calcipotriol against combination treatment with betamethasone valerate or diflucortolone valerate, once‐daily calcipotriol against combination treatment with fluocinonide acetonide or hydrocortisone, and twice‐daily calcitriol against combination treatment with diflucortolone valerate.

We reported elsewhere trials comparing vitamin D and combination therapy with vitamin D/potent corticosteroids for scalp psoriasis (Analysis 19.5).

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison summarises complex regimens for body psoriasis, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments (see Analysis 13.5 and Table 16). Using the combined end point, data were available for all 12 intervention‐comparator contrasts (Figure 4). Data were available from 2936 participants from 8 between‐patient trials and 1 within‐patient trial (Austad 1998). Trial duration varied between 2 and 12 weeks. One trial adequately concealed treatment allocation. As the interventions and comparators were highly variable and because two trials each contributed three pair‐wise contrasts, we did not pool the data.

Figure 4

---

---

Forest plot of comparison: 13 Vitamin D alone or in combination vs. other treatments: complex regimens, outcome: 13.5 Combined end point (IAGI/TSS/PASI/PAGI).

Seven intervention‐comparator contrasts found a significant difference between regimens. Six weeks' monotherapy with twice‐daily calcipotriol was less effective than clobetasol propionate (2 weeks) followed by 4 weeks' monotherapy with calcipotriol (SMD 0.60; 95% CI 0.18 to 1.02) (Austad 1998). A similar effect was evident when calcipotriol monotherapy (once daily for two weeks, then twice daily for four weeks) was compared with two weeks of treatment with calcipotriol (mornings) and fluocinonide acetonide (evenings), followed by 4 weeks' monotherapy with calcipotriol (SMD 0.66; 95% CI 0.01 to 1.32) (Wozel 2001). These benefits translated into improvements of 0.66 and 0.72 (i.e. around two‐thirds of a point) on a 6‐point IAGI scale.

The 12‐week trial by White 2006 (H)/White 2006 (P) compared 3 regimens; all included an initial phase in which participants applied once‐daily combination treatment with calcipotriol and betamethasone dipropionate for 4 weeks, but differed in their subsequent 8‐week maintenance phase. Maintenance therapy with twice‐daily placebo ointment was significantly less effective than maintenance with either calcipotriol ointment (SMD 0.27; 95% CI 0.12 to 0.41) or maintenance with calcipotriol on weekdays and combination treatment at weekends (SMD 0.51; 95% CI 0.37 to 0.66). When we directly compared the two regimens with active maintenance options, the alternating (weekday/weekend) approach was significantly more effective (SMD 0.26; 95% CI 0.11 to 0.40). The benefits of these 3 regimens equate to an improvement of 0.29, 0.56, and 0.28 of a point, respectively, on a 6‐point IAGI scale. These findings, based on the investigators' assessment of severity, were aligned with those of both the PASI assessment (Analysis 13.3) and the participants' assessment (PAGI) (Analysis 13.4).

Ortonne 2004 demonstrated that four weeks of treatment with a combination product of calcipotriol and betamethasone dipropionate followed by monotherapy with calcipotriol for four weeks was significantly more effective than monotherapy with tacalcitol (SMD 0.54; 95% CI 0.36 to 0.72) (0.59 on a 6‐point IAGI scale). Findings from the investigators' assessment (Analysis 13.1) were aligned with the PASI assessment (Analysis 13.3) and the participants' assessment (Analysis 13.4).

Kragballe 2004 compared 12 weeks of calcipotriol monotherapy with 2 complex regimens, both of which involved different sequences of combination therapy (calcipotriol/betamethasone dipropionate) and calcipotriol. When we directly compared these two complex regimens, 8 weeks of combination therapy followed by 4 weeks of once‐daily calcipotriol was significantly less effective than a routine involving 4 weeks of combination therapy followed by 8 weeks of once‐daily calcipotriol on weekdays and combination therapy at the weekend (SMD 0.24; 95% CI 0.08 to 0.40, equivalent to 0.27 (¼ of 1 point) on a 6‐point IAGI scale). This finding was based on the investigators' assessment (IAGI), which was very similar to the participants' (PAGI) assessment (Analysis 13.4). However, the PASI found a non‐significant trend in favour of the weekday/weekend regimen (Analysis 13.3).

In five intervention‐comparator contrasts, we found no significant difference. A single study reported two of the non‐significant intervention‐comparator contrasts: Kragballe 2004 compared 12 weeks of calcipotriol monotherapy with each of the 2 complex regimens described in the paragraph above. Compared with monotherapy, participants who applied once‐daily combination treatment (calcipotriol/betamethasone dipropionate) for 8 weeks followed by once‐daily calcipotriol for 4 weeks did not experience a significantly greater improvement (SMD ‐0.12; 95% CI ‐0.29 to 0.04). Similarly, the psoriasis of participants on monotherapy with twice‐daily calcipotriol did not improve significantly more or less than the disease in participants who applied once‐daily combination treatment (calcipotriol/betamethasone dipropionate) for 4 weeks followed by 8 weeks using once‐daily calcipotriol (weekdays) and once‐daily combination therapy (weekends) (SMD 0.13; 95% CI ‐0.04 to 0.29). These findings were based on the investigators' assessment (IAGI), which were consistent with those of the participants (Analysis 13.4) and with the PASI assessment (Analysis 13.3).

Similarly to Kragballe 2004, Saraceno 2007 compared 12 weeks of calcipotriol monotherapy with 4 weeks of combined therapy followed by 8 weeks of twice‐daily calcipotriol. The SMD for the combined end point (0.29; 95% CI ‐0.04 to 0.62) showed a non‐significant trend in favour of the complex regimen. Yang 2009 also explored how monotherapy with calcipotriol compared with combined vitamin D/corticosteroid treatment. A 6‐week course of monotherapy with calcipotriol was not significantly different to a sequential regimen involving halometasone (SMD 0.41; 95% CI ‐0.05 to 0.86). Lahfa 2003 found that the effect of clobetasol propionate combined with calcipotriol was similar to the effect of a regimen where the same corticosteroid was used in conjunction with calcitriol: SMD ‐0.19 (95% CI ‐0.54 to 0.16).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

This comparison included active‐controlled studies of psoriasis of the body that were at least 24 weeks in duration (see Analysis 14.5 and Table 17). One longer‐term placebo‐controlled trial of the body (Katz 1991a) and two long‐term scalp trials (Luger 2008; Poulin 2010) did not meet the inclusion criteria for this comparison, so we evaluated them elsewhere (Analyses 2, 19, and 18, respectively).

This comparison included one trial, Kragballe 2006, which was a between‐patient trial that compared 3 52‐week regimens with all treatments used once daily:

• first, combination treatment with calcipotriol and betamethasone dipropionate;

• second, alternating treatment with combination therapy for 4 weeks, then calcipotriol for 4 weeks; and

• third, combination therapy for 4 weeks, then calcipotriol for 48 weeks.

In total, 297 participants contributed data to the analysis. The only outcome measure from the trial that could be included in our review was the Investigator's Global Assessment of Disease Severity (scored from 0, absent, to 5, severe); therefore, results from this section are identical to those in Analysis 14.1. Data were unsuitable for pooling because the same participants contributed to more than one analysis. The adequacy of the concealment of treatment allocation was unclear in this trial.

Data from the combined end point showed there was no significant difference between the three long‐term regimens (Figure 5). One year's combination therapy was not significantly better than either the alternating regimen (SMD ‐0.09; 95% CI ‐0.36 to 0.18) or the regimen of treatment with 4 weeks of combination therapy followed by 48 weeks of calcipotriol (SMD ‐0.18; 95% CI ‐0.47 to 0.10). In both these comparisons, combination therapy achieved a larger absolute benefit, but the difference was not statistically significant. When we compared the alternating therapy with the regimen of 4 weeks' combination therapy followed by 48 weeks of calcipotriol, the SMD for the IAGI also indicated the 2 regimens were not statistically significantly different: ‐0.09 (95% CI ‐0.37 to 0.19).

Figure 5

---

---

Forest plot of comparison: 14 Vitamin D alone or in combination vs. other treatment: long term studies (>24wks), outcome: 14.5 Combined end point (IAGI/TSS/PASI/PAGI).

Findings were based on unpublished data supplied by the trial sponsor. Although these data were described as 'intention‐to treat', just 47% of the enrolled participants contributed data to the 52‐week assessment. The reasons why participants withdrew are summarised in Analysis 14.6, Analysis 14.7, and Analysis 14.8.

Analysis 15: Vitamin D analogues versus other treatment

This comparison incorporated all other vitamin D head‐to‐head comparisons of treatments for psoriasis of the body (excluding inverse psoriasis) that we had not already included (see Analysis 15.5 and Table 18). We included 12 intervention‐comparator contrasts, with data using the combined end point available for 11 of these contrasts. No effectiveness data were available for the comparison of calcipotriol and combined treatment with tazarotene gel plus mometasone furoate cream (Guenther 2000), but the trial did report data on withdrawal and adverse events. Thirteen between‐patient trials and 6 within‐patient trials provided data for the combined end point from 2364 participants. Trial duration ranged between 4 and 12 weeks. Three studies adequately concealed treatment allocation. In light of the pharmacological diversity of the comparators, we only pooled data within subgroups.

In three intervention‐comparator contrasts, vitamin D was significantly more effective than the comparator. Specifically, twice‐daily calcipotriol was more effective than coal tar polytherapy (SMD ‐0.59; 95% CI ‐0.87 to ‐0.31; I² statistic = 0%; 2 studies) and propylthiouracil cream (SMD ‐2.24; 95% CI ‐3.23 to ‐1.25; I² statistic = NA). On a 6‐point IAGI scale, these equate to improvements of 0.65 of a point and 2.46 points, respectively. Twice‐daily vitamin D was significantly more effective than once‐daily treatment (SMD ‐0.20; 95% CI ‐0.32 to ‐0.07; I² statistic = 0%; 3 studies), achieving an additional improvement of around 0.22 of a point on a 6‐point IAGI scale. This finding was based on dosing assessments of 2 vitamin D products: calcipotriol (SMD ‐0.19; 95% CI ‐0.37 to ‐0.02; I² statistic = 12%; 2 studies) and combination treatment with calcipotriol and betamethasone dipropionate (SMD ‐0.20; 95% CI ‐0.41 to 0.00; I² statistic = NA).

In the remaining eight intervention‐comparator contrasts, we found no significant difference between twice‐daily calcipotriol and the comparators. This finding held for the comparison against the following products: coal tar monotherapy (SMD ‐0.53; 95% CI ‐1.74 to 0.68; I² statistic = 91.1%; 3 studies); nicotinamide, either alone (SMD ‐0.09; 95% CI ‐0.49 to 0.31; I² statistic: NA) or in combination with calcipotriol (SMD 0.19; 95% CI ‐0.14 to 0.52; I² statistic = NA); betamethasone dipropionate ointment and salicylic acid (SMD ‐0.05; 95% CI ‐0.26 to 0.15; I² statistic = NA); tacrolimus ointment (SMD ‐0.55; 95% CI ‐1.28 to 0.17; I² statistic = 75.9%; 2 studies); tazarotene (SMD ‐0.10; 95% CI ‐0.35 to 0.16; I² statistic = 0%; 2 studies); and vitamin B12 cream (SMD ‐0.55; 95% CI ‐1.33 to 0.24; I² statistic = NA). The addition of occlusion to calcipotriol did not significantly improve the drug's effectiveness (SMD ‐0.18; 95% CI ‐2.04 to 1.68; I² statistic = 96.2%; 2 studies).

Two of the three trials comparing calcipotriol with coal tar monotherapy found a significant difference in favour of calcipotriol (De Simone 1993; Tham 1994); the other trial found a significant difference in favour of coal tar (Alora‐Palli 2010). This apparent contradiction may reflect different formulations of coal tar, treatment durations, or different baseline disease severity. Alora‐Palli 2010 also assessed effectiveness and quality of life over the six‐week post‐treatment period, finding that the coal tar group maintained their improvement significantly better than those treated with calcipotriol (see section (c) Quality of life measures).

The two occlusion trials used very different regimens and may be better considered separately. Hindsén 2006 (H) compared calcipotriol with occlusion (applied once weekly for 2 weeks) followed by 4 weeks off treatment against 6 weeks of calcipotriol monotherapy. This trial does not provide a simple assessment of occlusion, since, additionally, treatment is withdrawn. This parallel‐group trial contributed data from 209 participants with moderately severe disease and found monotherapy was significantly more effective at the 6‐week end point (SMD ‐1.11; 95% CI ‐1.40 to ‐0.81). In the other trial, 19 participants were treated for 8 weeks with twice‐daily calcipotriol on both sides of the body, and one side was randomised to receive overnight occlusion with polythene. Bourke 1993b found a significant difference in favour of occlusion (SMD 0.79; 95% CI 0.13 to 1.45); the trial did not explicitly state the severity of participants.

We reported elsewhere trials comparing vitamin D and other treatments for scalp psoriasis (Analysis 19.5) and inverse psoriasis (Analysis 17.5).

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of topical treatments for inverse or facial psoriasis (see Analysis 16.5 and Table 19). We found evidence on four treatments in this comparison: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus. We identified only one placebo‐controlled trial evaluating tacrolimus ointment (Lebwohl 2004), but the study did not report any effectiveness data suitable for this review. However, the study did contribute data on adverse events and withdrawal rates. We pooled only subtotals in this comparison.

Using the combined end point, data were available for three of the four topical treatments for inverse psoriasis. Two between‐patient trials contributed data from 122 participants. Gribetz 2004 was a eight‐week study that evaluated twice‐daily pimecrolimus cream. The four‐week trial by Kreuter 2006 (P) compared betamethasone valerate, calcipotriol, and pimecrolimus against placebo; all treatments were applied once daily. Treatment allocation was adequately concealed in one trial (Gribetz 2004). Therefore, we based the findings for each treatment on a single study.

The SMD for the combined end point found a statistically significant difference in favour of betamethasone valerate (SMD ‐2.83; 95% CI ‐3.79 to ‐1.88) and calcipotriol ointment (SMD ‐1.08; 95% CI ‐1.77 to ‐0.40). These equate to improvements of 3.25 points and 1.24 points on a 6‐point IAGI scale. Pimecrolimus cream was also significantly more effective than vehicle (SMD ‐0.86; 95% CI ‐1.30 to ‐0.41; I² statistic = 0%), equivalent to almost 1 point on a 6‐point IAGI. This result pooled data on twice‐daily applications for 8 weeks (SMD ‐1.07; 95% CI ‐1.69 to ‐0.45) and once‐daily pimecrolimus cream for 4 weeks (SMD ‐0.62; 95% CI ‐1.27 to 0.02). These findings suggest that different dosing schedules may influence efficacy.

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for inverse psoriasis, where we compared vitamin D with an active control (see Analysis 17.5 and Table 20). We identified five intervention‐comparator contrasts. Four treatments were compared with calcipotriol: once‐daily betamethasone valerate, combined treatment with calcipotriol and hydrocortisone, calcitriol, and pimecrolimus. Calcitriol was compared with tacrolimus.

Using the combined end point, data were available for all five intervention‐comparator contrasts. Four trials, varying in duration from 4 to 8 weeks, contributed data from 588 participants. Three studies were between‐patient trials, and one was a within‐patient study (Ortonne 2003). The adequacy of concealment of treatment allocation was unclear in all four trials.

When applied to sensitive areas of the body, calcipotriol was significantly less effective than three of the comparators. These included betamethasone valerate (SMD 2.02; 95% CI 1.20 to 2.84) (Kreuter 2006 (H)), combination treatment with calcipotriol and hydrocortisone (SMD 0.30; 95% CI 0.11 to 0.50) (Ortonne 2010), and calcitriol (SMD 0.61; 95% CI 0.28 to 0.94) (Ortonne 2003). On a 6‐point IAGI scale, the additional benefit equates to 2.22 points for betamethasone valerate, ⅓ of a point for combination treatment with hydrocortisone, and ⅔ of a point for calcitriol.

There was no significant difference between vitamin D and the topical calcineurin inhibitors (calcipotriol versus pimecrolimus: SMD ‐0.53 (95% CI ‐1.17 to 0.11; I² statistic = NA); calcitriol versus tacrolimus: SMD 0.42 (95% CI ‐0.15 to 0.98; I² statistic = NA)).

Analysis 18: Scalp psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of treatments for scalp psoriasis (see Analysis 18.5, Figure 6, and Table 21). We included evidence on 11 treatments in this comparison, with data from the combined end point available for all 11 treatments. Thirteen between‐patient trials and 1 within‐patient trial (Lepaw 1978) contributed data from 3011 participants. Trial duration ranged between two and eight weeks. Concealment of treatment allocation was adequate in one trial.

Figure 6

---

---

Forest plot of comparison: 18 Scalp psoriasis: placebo‐controlled trials, outcome: 18.5 Combined end point (IAGI/TSS/PASI/PAGI).

Two treatments were not significantly more effective than placebo. These were ciclopirox olamine shampoo (SMD ‐0.07; 95% CI ‐0.82 to 0.68; I² statistic = NA) and salicylic acid (SMD ‐0.86; 95% CI ‐1.79 to 0.06; I² statistic = NA). Of the nine treatments demonstrated to be significantly more effective than placebo, corticosteroids achieved the largest effects. Two very potent corticosteroids, amcinonide (SMD ‐1.42; 95% CI ‐1.80 to ‐1.04; I² statistic = NA) and clobetasol propionate (SMD ‐1.57; 95% CI ‐1.81 to ‐1.34; I² statistic = 43.3%; 4 studies), delivered benefits equating to almost 2 points (1.70 and 1.88 points, respectively) on a 6‐point IAGI (Investigator's Assessment of Global Improvement) scale. Betamethasone dipropionate combined with salicylic acid achieved a similar effect (SMD ‐1.48; 95% CI ‐2.50 to ‐0.47; I² statistic = NA), translating into 1.77 point improvement on the 6‐point IAGI over and above placebo.

When used as monotherapy, the 2 potent corticosteroids evaluated were both more effective than placebo: Betamethasone dipropionate (SMD ‐1.09; 95% CI ‐1.29 to ‐0.90; I² statistic = 0%; 2 studies) had a smaller effect than betamethasone valerate (SMD ‐1.40; 95% CI ‐1.75 to ‐1.05; I² statistic = NA), although the difference was not statistically significant. Calcipotriol had the smallest observed benefit (SMD ‐0.72; 95% CI ‐1.28 to ‐0.16; I² statistic = 69.2%; 2 studies). When calcipotriol was combined with betamethasone dipropionate, the pooled effect was larger than for calcipotriol used alone (SMD ‐0.97; 95% CI ‐1.61 to ‐0.32; I² statistic = 90.2%; 2 studies), or 1.16 points on a 6‐point improvement scale. Figure 6 shows that both trials found combination therapy to be significantly more effective, but Jemec 2008 (P) found a significantly larger effect (SMD ‐1.28; 95% CI ‐1.48 to ‐1.08) than Tyring 2010 (SMD ‐0.62; 95% CI ‐0.98 to ‐0.27). Participants in the two trials contributing data to the analysis of combination therapy differed in their ethnicity: Tyring 2010 recruited 177 people of Hispanic, Latino, Black, and African American ethnicity; in the trial by Jemec 2008 (P), over 96% of the 1505 enrollees were white. In addition, the trial by Jemec 2008 (P) included a larger proportion of people with severe or very severe disease (37% versus 20%). These factors may help explain the considerable level of heterogeneity (Higgins 2011) observed in the pooled effect for combination therapy.

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.5 and Table 22).

We identified six intervention‐comparator contrasts, and data on the combined end point were available for all these contrasts. All studies were parallel‐group in design (between‐patient). Twelve studies contributed data from 5413 participants. Trial duration ranged from 4 to 8 weeks in 11 studies, but duration was 52 weeks in one trial (Luger 2008). The adequacy of the concealment of treatment allocation was unclear in all 12 trials.

Based on the combined end point scores, monotherapy with vitamin D was consistently less effective than monotherapy with potent or very potent corticosteroids, and it was also significantly less effective than combination therapy with vitamin D and a potent steroid. Specifically, calcipotriol was significantly less effective than betamethasone dipropionate (SMD 0.48; 95% CI 0.32 to 0.64; I² statistic = 60.4%; 2 studies), betamethasone valerate (SMD 0.37; 95% CI 0.20 to 0.55; I² statistic = 0%; 2 studies), and clobetasol propionate (SMD 0.37; 95% CI 0.05 to 0.69; I² statistic = NA). On a 6‐point investigators' global assessment of improvement (IAGI) scale, these SMDs translate into 0.62, 0.48, and 0.48 points, respectively.

Combination treatment achieved a significantly greater benefit than either calcipotriol alone or betamethasone dipropionate alone. Calcipotriol was significantly less effective than combination treatment (SMD 0.64; 95% CI 0.44 to 0.84; I² statistic = 82.3%; 4 studies), with combined therapy delivering a benefit equivalent to 0.83 of a point on a 6‐point IAGI. Combination treatment was significantly more effective than monotherapy with betamethasone dipropionate (SMD ‐0.18; 95% CI ‐0.26 to ‐0.10; I² statistic = 0%; 3 studies), translating into a net benefit of 0.24 of a point on a 6‐point IAGI scale.

Compared with coal tar polytherapy, calcipotriol achieved a larger benefit, though this was not statistically significant at the 5% level (SMD ‐0.45; 95% CI ‐0.92 to 0.02; I² statistic = 89.8%; 3 studies).

Analysis 1: Vitamin D analogues versus placebo

We pooled withdrawal data from seven of the eight vitamin D analogues using a random‐effects risk difference (RD) metric (see Analysis 1.6, Analysis 1.7, and Analysis 1.8). There were no withdrawal data for the comparison of calcipotriol plus occlusion versus placebo.

There was no significant difference in the pooled rate of withdrawals from the trials for any reason (total withdrawals) (RD: ‐0.02; (95% CI ‐0.05 to 0.00; I² statistic = 51.4%). Rates of withdrawals due to adverse events were not statistically significantly different (RD ‐0.00; 95% CI ‐0.02 to 0.01; I² statistic = 36.4%), and neither were withdrawals due to treatment failure (RD ‐0.03; 95% CI ‐0.05 to 0.00; I² statistic = 81.7%). Individually, none of the eight vitamin D analogues were significantly different to placebo in any of the three withdrawal rates.

Analysis 2: Corticosteroid (potent) versus placebo

This comparison included 10 potent corticosteroids. No withdrawal data were available for diflorasone diacetate, and some withdrawal data were missing for fluticasone propionate, mometasone furoate, and betamethasone dipropionate. Where available, we pooled data using a random‐effects risk difference metric (see Analysis 2.6, Analysis 2.7, and Analysis 2.8). There was a small but statistically significant difference in favour of potent corticosteroids for withdrawals from the trials for any reason (total withdrawals): RD ‐0.14 (‐0.22 to ‐0.05; I² statistic = 81.5%). This finding was driven by two large trials of once‐daily betamethasone dipropionate (Fleming 2010 (P); Kaufmann 2002 (P)) and two small trials of betamethasone dipropionate used as maintenance (weekend pulse therapy) (Katz 1987a; Katz 1991a). In the other corticosteroids, the RD showed no statistically significant difference.

There was no significant difference in the pooled rate of withdrawals due to adverse events (RD ‐0.01; 95% CI ‐0.05 to 0.02; I² statistic = 60.9%). However, betamethasone dipropionate once daily was associated with statistically significantly lower rates of withdrawal due to adverse events than placebo (RD ‐0.07; 95% CI ‐0.11 to ‐0.02; I² statistic = NA).

For the rate of withdrawals due to treatment failure (Analysis 2.8), findings differed by treatment duration; therefore, we only pooled subgroup data. For short‐term treatments, there was no difference relative to placebo (RD 0.00; 95% CI ‐0.02 to 0.02; I² statistic = 0.0%). This is the pooled effect for the placebo comparisons with betamethasone valerate, budesonide, desonide, and hydrocortisone buteprate (data on withdrawals due to treatment failure for once‐daily and twice‐daily betamethasone dipropionate were not available). For maintenance treatment with betamethasone dipropionate, there was a statistically significant difference in favour of maintenance treatment when compared with placebo: RD ‐0.46; 95% CI ‐0.61 to ‐0.31; I² statistic = 0.0% (Katz 1987a; Katz 1991a).

Analysis 3: Corticosteroid (very potent) versus placebo

We identified withdrawal data for two of the three very potent corticosteroids in this comparison and pooled data using a random‐effects risk difference metric (see Analysis 3.6, Analysis 3.7, and Analysis 3.8). There were no withdrawal data available for halcinonide. There were no statistically significant differences between very potent corticosteroids and placebo for any type of withdrawal or for any individual treatment. For total withdrawals, the risk difference was ‐0.05 (95% CI ‐0.10 to 0.01; I² statistic = 83.7%). Differences in withdrawals due to adverse events (RD ‐0.00; 95% CI ‐0.01 to 0.01; I² statistic = 0.0%) and withdrawals due to treatment failure (RD ‐0.00; 95% CI ‐0.02 to 0.01); I² statistic = 13.5%) were also small and non‐significant.

Analysis 4: Dithranol versus placebo

We pooled withdrawal data on dithranol versus placebo using a random‐effects risk difference metric (see Analysis 4.6, Analysis 4.7, and Analysis 4.8). There were no statistically significant differences between dithranol and placebo for any type of withdrawal. For total withdrawals, the risk difference was 0.00 (95% CI ‐0.09 to 0.09; I² statistic = 0%). Differences in withdrawals due to adverse events (RD 0.00; 95% CI ‐0.05 to 0.05; I² statistic = 0%) and withdrawals due to treatment failure (RD 0.00; 95% CI ‐0.11 to 0.11; I² statistic = 0%) were also small and non‐significant.

Analysis 5: Vitamin D combination products versus placebo

We pooled data from trials of vitamin D combination products using a random‐effects risk difference (RD) metric (see Analysis 5.6, Analysis 5.7, and Analysis 5.8). In all three withdrawal measures, combined treatment ‐ whether used once or twice daily ‐ was significantly less likely to lead to withdrawal from the trial. This finding held for total withdrawals (RD ‐0.12; 95% CI ‐0.17 to ‐0.07; I² statistic = 59.3%), withdrawals due to adverse events (RD ‐0.07; 95% CI ‐0.11 to ‐0.04; I² statistic = 55.0%), and withdrawals due to treatment failure (RD ‐0.09; 95% CI ‐0.12 to ‐0.06; I² statistic = 0%).

Analysis 6: Other treatment versus placebo

We report findings separately for 22 of the 26 treatments considered; no withdrawal data were available for four placebo comparisons: those involving a herbal skin care product, topical sirolimus, topical tacrolimus, or coal tar. Data on total withdrawals were available for 22 treatments; for withdrawals due to adverse events or treatment failure, data were available for 18 treatments (see Analysis 6.6, Analysis 6.7, and Analysis 6.8).

We assessed withdrawal data using a random‐effects risk difference metric. Relative to placebo, tazarotene was associated with a significantly higher rate of withdrawal due to adverse events (RD 0.07; 95% CI 0.05 to 0.10) and Mahonia aquifolium was associated with a significantly lower rate of total withdrawal (i.e. withdrawal from the trial for any reason) (RD ‐0.23; 95% CI ‐0.32 to ‐0.14). No other treatment was significantly different from placebo on any of the three withdrawal measures assessed.

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

Seven of the eight vitamin D analogues that were compared with potent corticosteroids reported withdrawal data (withdrawal data on calcipotriol versus desoxymetasone were not available) (see Analysis 7.6, Analysis 7.7, and Analysis 7.8). We pooled withdrawal data on these seven treatment comparison pairs using a random‐effects risk difference (RD) metric.

Relative to potent corticosteroid, there was a statistically significant difference in total withdrawals in favour of corticosteroids: RD 0.02 (95% CI 0.00 to 0.03; I² statistic = 0%). Pooled rates of withdrawals due to adverse events or treatment failure were not statistically significantly different. Regarding individual vitamin D analogues, the only statistically significant differences in withdrawals relative to corticosteroid were for the comparison of calcipotriol against betamethasone dipropionate (total withdrawals: RD 0.03 (95% CI 0.01 to 0.06; I² statistic = 0%); withdrawals due to adverse events: RD 0.02 (95% CI 0.00 to 0.04; I² statistic = NA)).

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

We pooled withdrawal data on calcipotriol against clobetasol propionate using a random‐effects risk difference (RD) metric (see Analysis 8.6, Analysis 8.7, and Analysis 8.8). Relative to the very potent corticosteroid, we found no statistically significant difference for calcipotriol on any of the withdrawal assessments: total withdrawals, withdrawals due to adverse events, or withdrawals due to treatment failure.

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

We pooled withdrawal data on combination treatment (vitamin D and a corticosteroid) against monotherapy with a corticosteroid using a random‐effects risk difference (RD) metric (see Analysis 9.6, Analysis 9.7, and Analysis 9.8).

Relative to the corticosteroid, we found no statistically significant difference for combination treatment on total withdrawals, withdrawals due to adverse events, or withdrawals due to treatment failure. There were no data on the rate of withdrawals due to treatment failure for the comparison of combined treatment with calcipotriol and betamethasone dipropionate versus betamethasone dipropionate monotherapy.

Analysis 10: vitamin D alone or in combination versus dithranol

Withdrawal data were available for three intervention‐comparator pairs: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. We pooled these data using a random‐effects risk difference (RD) metric (see Analysis 10.6, Analysis 10.7, and Analysis 10.8). For the comparison of tacalcitol versus dithranol, there were no data available on withdrawals due to adverse events or treatment failure.

There was no statistically significant difference for total withdrawals, either for individual intervention‐comparator pairs or for the pooled effect (RD ‐0.02; 95% CI ‐0.06 to 0.01; I² statistic = 0%). The pooled risk difference showed a significant difference in favour of vitamin D for withdrawals due to adverse events (RD ‐0.03; 95% CI ‐0.06 to ‐0.00; I² statistic = 26.3%), but no significant difference in the rate of withdrawals due to treatment failure (RD ‐0.00; 95% CI ‐0.02 to 0.02; I² statistic = 0%).

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

Two of the three vitamin D analogues head‐to‐head comparisons reported withdrawal data (withdrawal data on calcipotriol versus tacalcitol were not available). We pooled withdrawal data on calcipotriol versus calcitriol, and calcipotriol versus maxacalcitol, using a random‐effects risk difference (RD) metric (see Analysis 11.6, Analysis 11.7, and Analysis 11.8). We found no statistically significant difference on any of the withdrawal assessments: total withdrawals, withdrawals due to adverse events, or withdrawals due to treatment failure. There was also no significant difference in the withdrawal rates for the individual contrasts, calcipotriol versus calcitriol, and calcipotriol versus maxacalcitol.

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

To simplify the analysis, we summarised withdrawal data by grouping the first 10 intervention‐comparator pairs in a single category: calcipotriol versus calcipotriol and corticosteroid. The remaining comparisons were calcitriol versus calcitriol and corticosteroid, and tacalcitol versus calcipotriol and corticosteroid. We pooled withdrawal data using a random‐effects risk difference (RD) metric (see Analysis 12.6, Analysis 12.7, and Analysis 12.8). There were no data on withdrawals due to treatment failure for the comparisons of combined therapy versus calcitriol or combined therapy versus tacalcitol.

In the comparison of calcipotriol against calcipotriol plus corticosteroid, total withdrawals were statistically significantly different in favour of polytherapy: RD 0.03 (95% CI 0.01 to 0.05; I² statistic = 13.3%). We based this on findings from 4922 participants in 13 trials. Similarly, a significant difference in favour of polytherapy was evident from the data on withdrawals due to adverse events: RD 0.02 (95% CI 0.01 to 0.03; I² statistic = 32.9%). However, we found no statistically significant difference for withdrawals due to treatment failure (RD 0.01; 95% CI ‐0.00 to 0.02; I² statistic = 0%). When either calcitriol or tacalcitol was compared with combination therapy, differences in total withdrawals and in withdrawals due to adverse events were not statistically significant.

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison covers complex regimens, defined here as treatment sequences that do not consist of a simple head‐to‐head comparison between two active treatments. We summarised withdrawals rates on twelve intervention‐comparator contrasts (see Analysis 13.6, Analysis 13.7, and Analysis 13.8). There were data on total withdrawal rates for all 12 contrasts, but data on withdrawals due to adverse events were missing for 3 contrasts, and there were no data on withdrawals due to treatment failure for 6 contrasts.

In the analysis of total withdrawals (Analysis 13.6), there was a significant difference in favour of the complex regimen for four contrasts:

• Participants were significantly more likely to withdraw from the trial after 12 weeks of calcipotriol than participants who received combination therapy (8 weeks) followed by calcipotriol (4 weeks) (RD 0.05; 95% CI 0.00 to 0.10).

• Trial withdrawal rates were also significantly higher in participants receiving 12 weeks of monotherapy with calcipotriol than those treated with combination therapy (4 weeks) followed by alternating calcipotriol on weekdays and combination therapy at the weekend (8 weeks) (RD 0.08; 95% CI 0.03 to 0.13).

• After an initial 4‐week phase with combination ointment, participants who received 8 weeks' maintenance with calcipotriol (RD 0.08; 95% CI 0.03 to 0.14) or 8 weeks' maintenance with an alternating calcipotriol/combined therapy routine (RD 0.11; 95% CI 0.06 to 0.17) were significantly less likely to withdraw than participants who used vehicle ointment for maintenance.

In the analysis of withdrawals due to adverse events (Analysis 13.7), there were no significant differences between vitamin D and any of the complex regimens.

In the analysis of withdrawals due to treatment failure (Analysis 13.8), there were significant differences in two of the six contrasts for which we had reported data. After 12 weeks of calcipotriol, participants were significantly more likely to withdraw from the trial because of treatment failure than those who received combination therapy followed by calcipotriol (RD 0.21; 95% CI 0.10 to 0.33). Withdrawals due to treatment failure were more likely in participants who received 8 weeks' monotherapy with tacalcitol than those who received combination therapy followed by calcipotriol (RD 0.05; 95% CI 0.02 to 0.08).

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

This comparison included active‐controlled trials of psoriasis of the body that were at least 24 weeks in duration (see Analysis 14.6, Analysis 14.7, and Analysis 14.8). There were three intervention‐comparator contrasts, all 52 weeks in duration and all reported by a single study (Kragballe 2006).

Participants received one of three long‐term treatments:

• once‐daily combined calcipotriol and betamethasone dipropionate;

• combination therapy (4 weeks) followed by calcipotriol (for 4 weeks), rotated over 52 weeks; or

• combination therapy for 4 weeks followed by calcipotriol for 48 weeks.

There were no significant differences between these three regimens in any of the withdrawal assessments: total withdrawals, withdrawals due to adverse events, or withdrawals due to treatment failure.

Analysis 15: Vitamin D analogues versus other treatment

This comparison compared vitamin D analogue with 12 other treatments. We assessed the results from the analyses of withdrawal data using a random‐effects risk difference (RD) metric, and we report these separately (see Analysis 15.6, Analysis 15.7, and Analysis 15.8). Withdrawal data were missing only in the head‐to‐head comparison of vitamin D (alone or in combination) versus calcipotriol with occlusion.

Of the remaining 11 intervention‐comparator contrasts, we found only 1 statistically significant difference in withdrawal rates: The rate of total withdrawals was significantly lower for calcipotriol than for tacrolimus (RD ‐0.13; 95% CI ‐0.25 to ‐0.01).

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of four topical treatments for inverse or facial psoriasis. Data on three withdrawal rates were available for all four treatments: the potent steroid betamethasone valerate; the vitamin D analogue calcipotriol; and two topical calcineurin inhibitors, pimecrolimus and tacrolimus (see Analysis 16.6, Analysis 16.7, and Analysis 16.8).

Relative to placebo, participants receiving topical tacrolimus were significantly less likely to withdraw from treatment for any reason (RD ‐0.17; 95% CI ‐0.30 to ‐0.03) or because of treatment failure (RD ‐0.11; 95% CI ‐0.19 to ‐0.02). No other treatment was significantly different from placebo on any of the three withdrawal measures assessed.

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for flexural or facial psoriasis, where we compared vitamin D with an active control (see Analysis 17.6, Analysis 17.7, and Analysis 17.8). We identified five intervention‐comparator contrasts, and no withdrawal data were missing. Four treatments were compared with calcipotriol: once‐daily betamethasone valerate, combined treatment with calcipotriol and hydrocortisone, calcitriol, and pimecrolimus. Calcitriol was compared with tacrolimus.

The rate of withdrawals due to adverse events was significantly higher in people treated with calcipotriol compared to those receiving combination therapy with calcipotriol and hydrocortisone for facial psoriasis (RD 0.06; 95% CI 0.02 to 0.11) and compared to those receiving calcitriol (RD 0.09; 95% CI 0.01 to 0.18). We found no other significant differences in withdrawal rates.

Analysis 18: Scalp psoriasis: placebo‐controlled trial

This comparison included placebo‐controlled trials of 11 topical treatments for scalp psoriasis (see Analysis 18.6, Analysis 18.7, and Analysis 18.8). All treatments had data on at least one withdrawal measure, which we assessed using a risk difference (RD) metric.

Relative to placebo, participants treated with betamethasone dipropionate were significantly less likely to withdraw from the trial for any reason (total withdrawals) (RD ‐0.14; 95% CI ‐0.21 to ‐0.06), because of adverse events (RD ‐0.04; 95% CI ‐0.08 to ‐0.00), or because of treatment failure (RD ‐0.10; 95% CI ‐0.16 to ‐0.05). Participants receiving combination therapy with calcipotriol and betamethasone dipropionate were significantly less likely to withdraw from the trial for any reason (total withdrawals) (RD ‐0.09; 95% CI ‐0.16 to ‐0.03) or because of treatment failure (RD ‐0.11; 95% CI ‐0.17 to ‐0.06). We found no other significant differences in withdrawal rates.

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.6, Analysis 19.7, and Analysis 19.8). We identified six intervention‐comparator contrasts. There were no data for withdrawal due to treatment failure for one of these contrasts (calcipotriol versus coal tar polytherapy). No other withdrawal data were missing.

Relative to combination treatment with calcipotriol and betamethasone dipropionate, calcipotriol monotherapy was associated with a significantly higher rate of total withdrawals (RD 0.11; 95% CI 0.05 to 0.18; I² statistic = 78.5%), withdrawals due to adverse events (RD 0.06; 95% CI 0.02 to 0.09; I² statistic = 78.9%) and withdrawals due to treatment failure (RD 0.05; 95% CI 0.01 to 0.10; I² statistic = 88.2%). Compared with betamethasone dipropionate, combination therapy was also significantly less likely to result in participant withdrawal due to treatment failure (RD ‐0.01; 95% CI ‐0.02 to ‐0.00; I² statistic = 0%).

Study participants treated with calcipotriol were significantly more likely to withdraw because of adverse events than participants receiving betamethasone valerate (RD 0.03; 95% CI 0.01 to 0.06; I² statistic = 0%) or coal tar polytherapy (RD 0.08; 95% CI 0.02 to 0.14; I² statistic = NA).

Analysis 1: Vitamin D analogues versus placebo

We pooled data on adverse events using a random‐effects risk difference metric (see Analysis 1.9 and Analysis 1.10). There were no adverse events data for the comparison of calcipotriol plus occlusion versus placebo. Therefore, data were available for seven of the eight vitamin D analogues evaluated against placebo

Of the seven vitamin D analogues evaluated against placebo, we found only one statistically significantly difference. Twice‐daily becocalcidiol was significantly more likely to cause local adverse events than placebo (RD 0.10; 95% CI 0.00 to 0.19). There were no significant differences for the pooled risk difference for either local adverse events (RD 0.00; 95% CI ‐0.01 to 0.02; I² statistic = 0%) or systemic adverse events (RD 0.00; 95% CI ‐0.01 to 0.01; I² statistic = 0%).

Analysis 2: Corticosteroid (potent) versus placebo

We pooled data on adverse events using a random‐effects risk difference metric (see Analysis 2.9 and Analysis 2.10). Data on local adverse events were available for 7 of the 10 corticosteroids; data on systemic adverse events were available only for betamethasone dipropionate twice daily, betamethasone dipropionate (maintenance), and mometasone furoate.

The pooled analysis of local adverse events found a significant difference in favour of corticosteroids (RD ‐0.04; 95% CI ‐0.08 to ‐0.00; I² statistic = 56.2%). Among the individual potent corticosteroids evaluated against placebo, we found only one statistically significantly difference. Once‐daily betamethasone dipropionate was less likely than placebo to be associated with local adverse events: RD ‐0.10 (95% CI ‐0.15 to ‐0.04; I² statistic = 2.5%). We found no significant difference in the pooled analyses of systemic adverse events data.

Analysis 3: Corticosteroid (very potent) versus placebo

We pooled data on adverse events using a random‐effects risk difference metric (see Analysis 3.9 and Analysis 3.10). The comparison with placebo found no statistically significant difference for any of the three very potent corticosteroids considered for either local adverse events or systemic adverse events.

Analysis 4: Dithranol versus placebo

We pooled data on adverse events using a random‐effects risk difference metric (see Analysis 4.9 and Analysis 4.10). The comparison with placebo found no statistically significant difference for pooled findings on local adverse events or systemic adverse events. In one study, a significant difference in favour of placebo was evident: RD 0.40 (95% CI 0.08 to 0.72; I² statistic = NA) (Volden 1992).

Analysis 5: Vitamin D combination products versus placebo

We pooled data on adverse events using a random‐effects risk difference metric (see Analysis 5.9 and Analysis 5.10). The rate of local adverse events was significantly lower for combination therapy compared with placebo (RD ‐0.05; 95% CI ‐0.08 to ‐0.02; I² statistic = 10.0%). There was no significant difference in the rate of systemic adverse events.

Analysis 6: Other treatment versus placebo

Our review reports findings separately for 22 of the 26 comparisons considered (see Analysis 6.9 and Analysis 6.10). No data on adverse events were available for placebo comparisons involving polymyxin B cream, topical sirolimus, topical tacrolimus, coal tar, or tazarotene. Data on local adverse events were available for 22 comparisons, but data on systemic adverse events were reported for just 9 of the 26 comparisons. We assessed data on adverse events using a random‐effects risk difference metric. The comparison with placebo found no statistically significant difference in the rate of either local or systemic adverse events for any treatment.

Analysis 7: Vitamin D analogues versus corticosteroid (potent)

Of the eight vitamin D analogues that were compared with potent corticosteroids, data on local adverse events were available for five comparisons, and data on systemic events were available for four comparisons (see Analysis 7.9 and Analysis 7.10). There were no data on adverse events for calcipotriol versus desoxymetasone, calcipotriol versus diflorasone diacetate, and calcitriol versus betamethasone valerate. In addition, systemic adverse events data were unavailable for calcipotriol versus fluocinonide. We pooled data on adverse effects using a random‐effects risk difference metric.

In the comparison of individual vitamin D analogues and potent corticosteroids, our review found one statistically significant difference in favour of potent corticosteroids. Pooled data from 3 trials with 1739 participants indicated that the rate of local adverse events was significantly higher in the calcipotriol group than in the group receiving betamethasone dipropionate (RD 0.07; 95% CI 0.04 to 0.09; I² statistic = 0%). This finding drove the pooled (class) result for local adverse events (RD 0.07; 95% CI 0.02 to 0.11; I² statistic = 82.4%). Our review found no other statistically significant differences for local or systemic adverse events in this comparison.

Analysis 8: Vitamin D analogues versus corticosteroid (very potent)

We pooled withdrawal data on adverse effects using a random‐effects risk difference metric (see Analysis 8.9 and Analysis 8.10). There was no statistically significantly difference between calcipotriol and clobetasol propionate, either for local adverse events (RD ‐0.05; 95% CI ‐0.18 to 0.08) or systemic events (RD ‐0.05; 95% CI ‐0.18 to 0.08).

Analysis 9: Vitamin D combined with corticosteroid versus corticosteroid

We pooled withdrawal data on combination treatment (vitamin D and a corticosteroid) against monotherapy with a corticosteroid using a random‐effects risk difference metric (see Analysis 9.9 and Analysis 9.10). No adverse events data were available for the comparison of combination (calcipotriol/clobetasol propionate) therapy and clobetasol propionate.

The pooled analysis found no statistically significant difference in adverse event rates between combination (calcipotriol/betamethasone dipropionate) therapy and betamethasone dipropionate. There was no significant difference in local adverse events for combination (calcipotriol/betamethasone) therapy and clobetasol propionate, and there were no data on systemic adverse events for this comparison.

Analysis 10: Vitamin D alone or in combination versus dithranol

Adverse events data were available for three intervention‐comparator pairs: calcipotriol versus dithranol, calcitriol versus dithranol, and tacalcitol versus dithranol. We pooled these data using a random‐effects risk difference metric (see Analysis 10.9 and Analysis 10.10).

Vitamin D was statistically significantly less likely to cause local adverse events (RD ‐0.32; 95% CI ‐0.43 to ‐0.20; I² statistic = 84.5%), although substantial heterogeneity was evident in the pooled statistic (Higgins 2011). This finding also held for each of the three intervention‐comparator pairs, with the effect size ranging from ‐0.25 (calcipotriol versus dithranol) to ‐0.67 (calcitriol versus dithranol). However, the analysis of systemic adverse events found no statistically significant differences.

Analysis 11: Vitamin D alone or in combination versus other vitamin D analogue

We reported local adverse events data on two of the vitamin D analogues head‐to‐head comparisons (see Analysis 11.9 and Analysis 11.10). Systemic adverse events data were not available for any of the three comparisons. We pooled data on adverse effects using a random‐effects risk difference metric. Overall, our review found no statistically significant difference on either local or systemic adverse events when comparing pooled vitamin D analogues head‐to‐head. However, in the comparison of calcipotriol against calcitriol, the rate of local adverse events was significantly lower for calcitriol: RD 0.07 (95% CI 0.01 to 0.14; I² statistic = NA).

Analysis 12: Vitamin D alone or in combination versus vitamin D + corticosteroid

To simplify the comparison, we summarised adverse events data by grouping the first 10 intervention‐comparator pairs in a single category: calcipotriol versus calcipotriol and corticosteroid. The remaining comparisons were 1) calcitriol versus calcitriol and 2) corticosteroid and tacalcitol versus calcipotriol and corticosteroid. There were no data on systemic adverse events for the comparison of tacalcitol against calcipotriol and corticosteroid.

We pooled data on adverse effects using a random‐effects risk difference (RD) metric (see Analysis 12.9 and Analysis 12.10). In the local adverse events comparison of vitamin D against vitamin D plus corticosteroid, our review found a statistically significantly difference in favour of combination therapy: RD 0.06 (95% CI 0.05 to 0.08; I² statistic = 4.1%). This finding also held for each individual intervention‐comparator pair. Our analysis found no statistically significantly difference in the rates of systemic adverse events.

Analysis 13: Vitamin D alone or in combination versus other treatments: complex regimens

This comparison included 12 intervention‐comparator contrasts (see Analysis 13.9 and Analysis 13.10). Data on local adverse events were available for 11 contrasts, but there were no data on the comparison of calcipotriol (12 weeks) versus combination therapy (4 weeks) followed by calcipotriol (8 weeks). No trial in this comparison reported data on systemic adverse events.

In 4 of the 11 contrasts for which data were available, the complex regimen was associated with a significantly lower rate of local adverse events than when the vitamin D analogue was used alone. Twelve weeks of calcipotriol monotherapy was associated with significantly higher rates of local adverse events than two complex regimen comparators. Specifically, the rate of local adverse events was higher for calcipotriol monotherapy than for a regimen consisting of 8 weeks of combination therapy followed by 4 weeks of once‐daily calcipotriol (RD 0.11; 95% CI 0.06 to 0.17). Calcipotriol monotherapy was also more likely to be associated with local adverse events than 4 weeks of combination therapy followed by 8 weeks of once‐daily calcipotriol on weekdays and combination therapy at the weekend (RD 0.11; 95% CI 0.05 to 0.17) (Kragballe 2004). When we compared the two complex regimens directly, the rates of cutaneous adverse events were not significantly different.

Relative to calcipotriol, combination therapy with halometasone and calcipotriol had a significantly lower rate of adverse events of the skin (RD 0.26; 95% CI 0.07 to 0.45). An 8‐week course of tacalcitol had significantly higher rates of adverse events than a routine of 4 weeks' combined (calcipotriol/betamethasone dipropionate) therapy followed by 4 weeks' monotherapy with calcipotriol (RD 0.06; 95% CI 0.01 to 0.11).

There was no significant difference in the rate of local adverse events in any of the remaining seven intervention‐comparator contrasts.

Analysis 14: Vitamin D alone or in combination versus other treatment: long‐term studies (> 24 weeks)

This comparison included active‐controlled studies of psoriasis of the body that were at least 24 weeks in duration (see Analysis 14.9 and Analysis 14.10). No data on systemic adverse events were reported. We included one trial in this comparison: Kragballe 2006 compared 3 52‐week regimens with all treatments used once daily.

Combination therapy with calcipotriol and betamethasone dipropionate was associated with significantly lower rates of adverse events than either of the two 'complex' long‐term regimens:

• alternating treatment with combination therapy for 4 weeks, then calcipotriol for 4 weeks (RD ‐0.08; 95% CI ‐0.17 to ‐0.00); and

• combination therapy for 4 weeks, then calcipotriol for 48 weeks (RD ‐0.16; 95% CI ‐0.25 to ‐0.08).

The difference between the two 'complex' regimens was not statistically significant at the 5% level (RD ‐0.08; 95% CI ‐0.17 to 0.01), although there was a trend in favour of alternating therapy.

Analysis 15: vitamin D analogues versus other treatment

A vitamin D analogue was compared with 12 other treatments (see Analysis 15.9 and Analysis 15.10). Data on local adverse events were available for 10 intervention‐comparator contrasts, and systemic adverse events data were available for 8 intervention‐comparator contrasts.

We assessed results from the analyses of adverse events data using a random‐effects risk difference metric and report these separately. For local adverse events, three intervention‐comparator contrasts were significantly different. These were the comparison of calcipotriol against calcipotriol + nicotinamide (RD ‐0.17; 95% CI ‐0.30 to ‐0.03), calcipotriol versus corticosteroid + salicylic acid (RD 0.09; 95% CI 0.02 to 0.15), and calcipotriol versus tacrolimus ointment (RD ‐0.19; 95% CI ‐0.37 to ‐0.01). No trial that reported the rate of systemic adverse events found a significant difference in this outcome.

Analysis 16: Flexural/facial psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of four topical treatments for inverse or facial psoriasis. Data on local adverse events were available for all four treatments, but data on systemic events were reported only for one treatment, i.e. the comparison of tacrolimus against placebo (see Analysis 16.9 and Analysis 16.10).

Relative to placebo, participants receiving topical tacrolimus were significantly less likely to report local adverse events (RD ‐0.17; 95% CI ‐0.30 to ‐0.03), but there was no significant difference in the rate of systemic events. No other treatment was significantly different from placebo.

Analysis 17: Flexural/facial psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for flexural or facial psoriasis, where vitamin D was compared with an active control. We identified five intervention‐comparator contrasts (see Analysis 17.9 and Analysis 17.10). Local adverse events data were missing for the comparison of calcitriol and tacrolimus. There were no data on systemic events for any of the five contrasts.

In people treated with calcipotriol, the rate of local adverse events was significantly higher compared to those receiving combination therapy with calcipotriol and hydrocortisone for facial psoriasis (RD 0.15; 95% CI 0.08 to 0.23) and also compared to those receiving calcitriol (RD 0.09; 95% CI 0.02 to 0.17). These findings aligned with the withdrawal rates for adverse events (Analysis 17.7). We found no other significant differences in adverse events rates.

Analysis 18: Scalp psoriasis: placebo‐controlled trials

This comparison included placebo‐controlled trials of 11 topical treatments for scalp psoriasis (see Analysis 18.9 and Analysis 18.10). There were no adverse events data for two treatments, betamethasone valerate and amcinonide. Data on systemic events were available for four treatments.

Relative to placebo, participants treated with betamethasone dipropionate were significantly less likely to suffer local adverse events (RD ‐0.07; 95% CI ‐0.13 to ‐0.01; I² statistic = 0%). These findings aligned with the withdrawal rates for adverse events (Analysis 18.7). We found no other significant differences in adverse events rates.

Analysis 19: Scalp psoriasis: vitamin D alone or in combination versus other treatment

This comparison included head‐to‐head trials of treatments for scalp psoriasis in which one of the interventions was a vitamin D product (used either as monotherapy or in combination with another product) (see Analysis 19.9 and Analysis 19.10). We identified six intervention‐comparator contrasts, and there were no adverse events data.

Calcipotriol monotherapy was associated with a significantly higher rate of local adverse events than five of the comparator treatments; these included betamethasone dipropionate (RD 0.07; 95% CI 0.04 to 0.11; I² statistic = 0%), betamethasone valerate (RD 0.17; 95% CI 0.01 to 0.33; I² statistic = 76.5%), and clobetasol propionate (RD 0.19; 95% CI 0.10 to 0.28; I² statistic = 0%). Relative to calcipotriol, combined therapy with calcipotriol and betamethasone dipropionate (RD 0.09; 95% CI 0.06 to 0.12; I² statistic = 28.1%), and coal tar polytherapy (RD 0.24; 95% CI 0.15 to 0.33; I² statistic = NA), were also associated with significantly lower rates of adverse events. Compared with betamethasone dipropionate monotherapy, the rate of adverse events for combined therapy with calcipotriol and betamethasone dipropionate was not significantly different (RD ‐0.00; 95% CI ‐0.02 to 0.01; I² statistic = NA). There were no significant differences in the rates of systemic adverse events.