Scolaris Content Display Scolaris Content Display

Compresión neumática intermitente para el tratamiento de las úlceras venosas de la pierna

Contraer todo Desplegar todo

Resumen

Antecedentes

La compresión neumática intermitente (CNI) es un método mecánico de aplicación de compresión a las extremidades tumefactas que se puede utilizar para el tratamiento de las úlceras venosas de la pierna y el edema de las extremidades debido a un linfedema.

Objetivos

Determinar si la CNI aumenta la cicatrización de las úlceras venosas de la pierna. Determinar los efectos de la CNI sobre la calidad de vida relacionada con la salud de los pacientes con úlceras venosas de la pierna.

Métodos de búsqueda

En abril de 2014, para esta tercera actualización, se realizaron búsquedas en el Registro especializado del Grupo Cochrane de Heridas (Cochrane Wounds Group); en el Registro Cochrane central de ensayos controlados (Cochrane Central Register of Controlled Trials, CENTRAL)(The Cochrane Library); Ovid MEDLINE; Ovid MEDLINE (In‐Process & Other Non‐Indexed Citations); Ovid EMBASE; y EBSCO CINAHL.

Criterios de selección

Se incluyeron los ensayos controlados aleatorizados (ECA) que compararon los efectos de la CNI con control (CNI simulada o ninguna CNI) o que realizaron comparaciones entre regímenes de tratamiento con CNI para las úlceras venosas.

Obtención y análisis de los datos

Dos autores de la revisión revisaron los títulos y los resúmenes y acordaron los estudios completos que se debían recuperar. Un autor de la revisión extrajo los datos y evaluó los estudios en cuanto al riesgo de sesgo, lo que fue verificado por un segundo autor de la revisión.

Resultados principales

Se identificaron nueve ensayos controlados aleatorizados (con 489 personas en total). Sólo un ensayo tuvo un bajo riesgo de sesgo en general al haber informado de una adecuada asignación al azar, ocultación de la asignación y una evaluación cegada de los desenlaces.
En un ensayo (80 personas), hubo un número mayor de úlceras cicatrizadas con la CNI que con los apósitos (62% versus 28%; p = 0,002). Cinco ensayos compararon la CNI más compresión con compresión sola. Dos de ellos (97 personas) encontraron una mayor cicatrización de las úlceras con la CNI más compresión en comparación con la compresión sola. Los tres ensayos restantes (122 personas) no encontraron evidencia de un efecto beneficioso de la CNI más compresión en comparación con la compresión sola.

Dos ensayos (86 personas) no encontraron diferencias entre la CNI (sin compresión adicional) y las vendas de compresión solas.

Un ensayo (104 personas) comparó diferentes formas de administración de la CNI y encontró que con la CNI rápida cicatrizaron más úlceras que con la CNI lenta (86% versus 61%).

Conclusiones de los autores

La CNI puede aumentar la cicatrización en comparación con ninguna compresión. No está claro si se puede usar en lugar de las vendas de compresión. Hay alguna evidencia limitada de que la CNI puede mejorar la cicatrización cuando se agrega a las vendas de compresión. En un ensayo, la CNI rápida fue mejor que la CNI lenta. Se necesitan ensayos adicionales para determinar la fiabilidad de la evidencia actual, qué pacientes se pueden beneficiar con la CNI además de las vendas de compresión y el régimen de tratamiento óptimo.

Resumen en términos sencillos

Compresión neumática intermitente para el tratamiento de las úlceras venosas de la pierna.

La obstrucción o el mal funcionamiento de las venas de la pierna pueden causar úlceras venosas (heridas abiertas). La compresión, mediante vendas o medias, puede ayudar a cicatrizar las úlceras. Sin embargo, no siempre funcionan, y algunas personas no son capaces o no están dispuestas a utilizarlas. La compresión neumática intermitente (CNI) utiliza una bomba de aire para inflar y desinflar una bolsa hermética envuelta alrededor de la pierna. Esta técnica también se utiliza para evitar que se desarrollen coágulos de sangre durante la cirugía. Sin embargo, la revisión de los ensayos encontró evidencia contradictoria acerca de si la CNI es mejor que las vendas y las medias de compresión. La compresión neumática intermitente (CNI) es mejor para la cicatrización de las úlceras de la pierna que ninguna compresión. Algunos estudios indican que la CNI podría ser un complemento beneficioso a las vendas en el caso de algunas úlceras, aunque estos estudios podrían estar sesgados. La administración del tratamiento con CNI de una manera rápida, inflando y desinflando el dispositivo de CNI más rápidamente, dio lugar a que más úlceras cicatrizaran en comparación con un régimen de desinflado más lento.

Authors' conclusions

Implications for practice

IPC improved healing in a poor quality trial, when compared with dressings alone, and this reinforces the finding that continuous compression is better than no compression for venous ulcer healing. There is some evidence that the addition of IPC to standard compression may be beneficial. This, however, is based on a small number of small to medium sized studies, which are open to bias. These studies also used different IPC regimens, and there is no evidence as to how long IPC is indicated for, and no evidence as to the optimum IPC regimen. Having said this, there is some evidence that more ulcers healed when a 'fast' IPC therapy was used, which delivers IPC inflation and deflation of the compression quicker, than with a 'slow' IPC therapy. There is also some evidence of lower pain scores using IPC over compression.

Implications for research

Larger, well designed RCTs comparing IPC in both hospital and home settings are indicated. Questions that remain to be answered are:

  1. Is IPC beneficial in people with venous ulcers who cannot tolerate compression bandages or hosiery ?

  2. Is IPC useful as an adjuvant to compression therapy (bandages or hosiery) ?

  3. Is IPC an effective alternative to compression bandages / hosiery for some patients who would prefer not to wear compression 24 hours a day ?

  4. If IPC is beneficial in any of the above circumstances, what is the optimum IPC regimen?

Studies must measure cost data as well as patient outcomes such as quality of life, pain and acceptability of IPC treatment.

Background

Venous leg ulcers are a common chronic complaint affecting some 3% of the population over the age of 65 years (Callam 1985). They are caused by chronic venous insufficiency, secondary to leg veins being blocked or incompetent. Compression, using roller bandages or hosiery, has been shown to improve ulcer healing rates compared with no compression (Cullum 1997). Bandages or hosiery heal around 70% of venous leg ulcers after 3 to 6 months of treatment (Cullum 1997), but a substantial proportion of patients are not helped by compression bandaging, or are unwilling or unable to wear it. Alternative methods of treating venous leg ulcer patients are required.

Intermittent pneumatic compression (IPC) uses an air pump to periodically inflate/ deflate bladders incorporated into sleeves which envelop legs or arms. There are a number of ways of applying intermittent pneumatic compression, using single or multiple chambers / bladders, or by using different types of pumps and compression cycles, for example, variation of inflation and deflation times. IPC has been used to prevent the development of blood clots during long periods of rest, for example, during surgery. It has also been used to treat limb swelling, for example, lymphoedema, and venous leg ulcers. The purpose of this review is to examine the effectiveness of IPC as a treatment of venous ulcers, and to analyse the evidence for the effectiveness of different IPC regimens.

Objectives

To review the best available evidence for the effects of IPC on venous leg ulcer healing and quality of life.

Specific questions addressed by the review are:

  1. does IPC increase the healing of venous leg ulcers?

  2. what are the relative effects of different IPC regimens?

  3. does IPC affect pain or quality of life?

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs) making the following comparisons:

  • intermittent pneumatic compression (IPC) compared with a control treatment, namely sham IPC or standard care;

  • intermittent pneumatic compression (IPC) plus compression compared with compression alone;

  • one IPC regimen compared with an alternative IPC regimen.

Studies performed in the home, in hospital or in any other setting were considered for inclusion in the review.

Types of participants

People of any age with venous leg ulcers, irrespective of method of diagnosis of venous insufficiency, and regardless of level of mobility.

Types of interventions

IPC applied to the leg. All types of devices were included, for example, below knee or thigh length, single or multiple compartment devices, irrespective of the duration or frequency of treatment.

Types of outcome measures

Primary outcomes

  • time to complete healing of ulcers;

  • proportion of ulcers healing in the trial period;

  • rate of reduction of ulcer area (expressed either as a percentage or an actual area reduction per unit time).

Secondary outcomes

  • pain and quality of life as measured by validated scales

Change in limb volume was not included as an outcome measure as there is only circumstantial evidence that reduction of limb swelling leads to ulcer healing.

Search methods for identification of studies

Electronic searches

For an outline of the search methods used in second update of this review see Appendix 1.

In April 2014, for this third update, we searched the following electronic databases:

  • The Cochrane Wounds Group Specialised Register (searched 01 April 2014);

  • The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2014, Issue 1);

  • The Health Technology Assessment Database (HTA) (The Cochrane Library 2014, Issue 1);

  • Ovid MEDLINE (1946 to March Week 3 2014);

  • Ovid MEDLINE (In‐Process & Other Non‐Indexed Citations, March 31, 2014);

  • Ovid EMBASE (1974 to 2014 March 31);

  • EBSCO CINAHL (1982 to 01 April 2014)

We used the following strategy to search the Cochrane Central Register of Controlled Trials (CENTRAL):
#1 MeSH descriptor: [Intermittent Pneumatic Compression Devices] explode all trees
#2 (pneumatic near/10 compression*):ti,ab,kw
#3 (compression near/10 device*):ti,ab,kw
#4 (flowtron or IPC):ti,ab,kw
#5 (sequential near/10 compression*):ti,ab,kw
#6 #1 or #2 or #3 or #4 or #5
#7 MeSH descriptor: [Leg Ulcer] explode all trees
#8 MeSH descriptor: [Varicose Ulcer] explode all trees
#9 ((varicose next ulcer*) or (venous next ulcer*) or (leg next ulcer*) or (stasis next ulcer*) or (crural next ulcer*) or “ulcus cruris” or “ulcer cruris”):ti,ab,kw
#10 #7 or #8 or #9
#11 #6 and #10

The search strategies for Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL can be found in Appendix 2, Appendix 3 and Appendix 4 respectively. We combined the Ovid MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximising version (2008 revision); Ovid format (Lefebvre 2011). We combined the EMBASE search with the Ovid EMBASE filter developed by the UK Cochrane Centre (Lefebvre 2011). There were no restrictions on the basis of date or language of publication and translations of all relevant non‐English language papers were obtained using local resources.

Searching other resources

We searched the bibliographies of all retrieved and relevant publications identified by these strategies for further studies.

Data collection and analysis

Selection of studies

Two (of four; RM, EAN, AH and KV) review authors independently examined the titles and abstracts of citations generated by the search to identify those likely to meet the inclusion criteria and retrieved these in full. We (RM, EAN, AH or KV) checked full versions of articles to identify those that met the inclusion criteria.

Data extraction and management

We extracted and summarised details of eligible trials using a data extraction sheet. Studies published in duplicate were included only once. One review author undertook data extraction (EAN or AH) and this was checked by a second review author (EAN, RM or KV). One review author undertook the risk of bias assessment (KT or AH) and this was checked by a second review author (EAN).

Where available, data were collected on the following (see Characteristics of included studies for detail):

  • author, title, date of study and publication;

  • sample size;

  • participant inclusion and exclusion criteria;

  • care setting;

  • baseline variables, for example age, sex, diagnosis, co‐morbidity, baseline risk, baseline comparability of treatment groups for important variables;

  • description of interventions;

  • numbers of participants ‐ both randomised and analysed;

  • description of any co‐interventions;

  • follow‐up period, extent of loss to follow up;

  • results;

  • outcome measures;

  • adverse events;

  • allocation concealment

  • use of intention to treat analysis

  • blinded outcome assessment;

  • and trialists' conclusions.

Assessment of risk of bias in included studies

For the update of this review one review author (KT or AH) assessed each included study using the Cochrane Collaboration tool for assessing risk of bias and this was checked by a second review author (EAN) (Higgins 2011).

This tool addresses seven specific domains, namely sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and other bias (e.g. extreme baseline imbalance) (see Appendix 5 for details of criteria on which the judgement is based). We assessed blinding and completeness of outcome data for each outcome separately. We completed a 'Risk of bias' table for each eligible study and discussed any disagreement to achieve a consensus.

We presented an assessment of risk of bias using a 'Risk of bias' summary figure (Figure 1), which presents all of the judgements in a cross‐tabulation of study by entry. This display of internal validity indicates the weight the reader may give the results of each study.


Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Data synthesis

Data were entered into and analysed using Cochrane RevMan software. Results are presented with 95% confidence intervals (CI). Estimates for dichotomous outcomes (e.g. number of ulcers healed ‐ yes or no) were reported as risk ratio (RR). Continuous data (e.g. total area healed, or changes in volume of ulcers) were converted to the standardised mean difference (or a weighted mean difference, when plausible) and overall effect size (with 95% CI) calculated. Time to wound healing and time to return to work were analysed as survival (time to event) data, using the appropriate analytical method (as per the Cochrane Reviewers' Handbook version 5.0.2). Methods of synthesising the studies depend on their quality, design and heterogeneity. Both clinical and statistical heterogeneity were explored. Clinical heterogeneity was explored by examining potentially influential factors such as parameters of IPC used, care setting, co‐interventions such as the use of compression therapy. In the absence of clinical and statistical heterogeneity a fixed effect model was applied to pool data. In the presence of statistical heterogeneity (as estimated by the I2 where values greater than 50% indicate the presence of substantial heterogeneity) (Higgins 2003) a random effects model was applied for meta‐analysis. Where synthesis is inappropriate we undertook a narrative overview.

Results

Description of studies

The initial search yielded eight potentially relevant citations to seven studies. Two were controlled clinical trials in venous leg ulcer treatment (Belcaro 1986; Pekenmaki 1987), one mentioned IPC as an additional treatment after a trial of surgical techniques (Sottiurai 1990) resulting in three studies which did not meet the inclusion criteria for the review (see Characteristics of excluded studies). Four studies were included (Coleridge Smith 1989; McCulloch 1994; Rowland 2000; Schuler 1996).

For the updates of this review, searches yielded 55 citations. These were considered by two review authors independently and resulted in five additional trials meeting the inclusion criteria (Alvarez 2012; Dolibog 2013; Kumar 2002; Nikolovska 2002; Nikolovska 2005). Nikolovska 2002 was identified from searching the reference list of the included studies. Alvarez 2012 was identified as an ongoing study in the second update, and included in the third update after personal correspondence with the author identified sufficient information to include the study.

One study was classified as 'awaiting assessment' pending full text retrieval (Kussmann 2005), we have been unable to obtain a copy of this paper and therefore have added the citation to the Characteristics of excluded studies table along with the other 25 citations to 24 studies which did not meet the inclusion criteria for this review.

In total, nine RCTs were included in the updated review. The studies were conducted in outpatient department / home environments (7 trials) and outpatient /inpatient (2 trials). The studies are described below in relation to the treatment comparisons.

IPC plus dressings compared with dressings alone (1 trial, 80 participants)

One trial evaluated whether adding IPC to a dressing regimen increased healing rates. (Nikolovska 2002). Nikolovska randomised 80 people (by coin toss) to a hydrocolloid dressings and either standard care (no compression) or IPC: multi‐chamber, sequential device with 50 mmHg at ankle and 40 mmHg at thigh, with 3 minute inflation time, 30 seconds at full pressure and 60 second deflation time; 1 hour, five days a week, for up to six months.

IPC plus continuous compression compared with continuous compression alone (5 trials, 219 participants)

Five trials evaluated the effect of adding IPC to a standard regimen of compression bandages or stockings (Alvarez 2012; Coleridge Smith 1989; Kumar 2002; McCulloch 1994; Schuler 1996).

Alvarez 2012 compared IPC plus standard compression therapy with compression therapy alone in 52 patients. Compression therapy consisted of a non‐adherent primary wound dressing plus a 4‐layer compression bandage. IPC was performed using a 4‐chamber pneumatic leg sleeve and gradient, sequential pump. All pumps were calibrated to a pressure setting of 40‐50mmHg, with treatment lasting one hour, twice daily. Inflation and deflation time is not reported. Wounds were measured using a 3 mega pixel digital camera and photo‐digital planimetry software. Randomisation was by a computer random number generator.

Coleridge Smith 1989, in a trial of 45 people, assessed the addition of IPC to a standard regimen in which ulcers were dressed identically and legs compressed with stockings designed to exert 30 to 40 mmHg graduated compression. A sequential compression device was used to deliver IPC (number of cells is not stated) with higher pressures exerted distally and reducing up the leg. Randomisation was determined by coin tossing. Pre‐trial assessment included measurement of ankle brachial pressure index (ABPI), Duplex ultrasound and plethysmographic assessment of venous insufficiency. The 21 participants in the IPC group inserted the ulcerated, stockinged limb into the IPC foot‐sleeve daily for up to 4 hours. All patients in the study were advised to elevate their legs while resting, and to avoid standing. Dressings were changed weekly. Ulcer area was measured, by tracing, at the start and at 2 weekly intervals during dressing changes, but the baseline comparability of the median ulcer area was poor, the IPC group having larger ulcers than the control group.

Kumar 2002 compared IPC plus four‐layer bandaging with four‐layer bandaging alone in 47 people. Allocation was by coin toss and data collectors were unblinded. The IPC was a below‐knee, single chambered device, with 90 second inflation and 90 second deflation, at 60 mmHg, used for 60 minutes, twice a day, for up to 4 months. This trial recruited 47 people (54 ulcers) but only reported outcomes on the 41 people (48 ulcers) who completed the trial. Kumar 2002 randomised people but included and reported on the number of ulcers healed, some participants had more than one ulcer. This raises concern of a unit of analysis error.

McCulloch 1994 compared IPC plus Unna's Boots with Unna's Boots alone on 22 patients with venous leg ulcers. IPC was delivered twice a week using a single cell device which compressed the limb uniformly, with care delivered at home. Fifty mmHg of pressure was applied for 90 seconds followed by 30 seconds of relaxation after which legs were dressed in Unna's Boots. Co‐interventions were the same in both groups. "Randomisation" was by case note number. No objective assessment of venous insufficiency was reported. Ulcer area was measured by tracing at dressing changes.

Schuler 1996 compared IPC ("HomeRx" sequential gradient compression device) plus elastic stockings with Unna's Boot therapy alone in 53 participants. The ulcers of people in both groups were dressed with hydrocolloid dressings. All patients were advised to elevate their legs twice daily. The experimental group was supplied with elastic stockings, graded to deliver 30 mmHg at the ankle. These stockings were removed for IPC therapy ("HomeRx") for 1 hour/am and 2 hours/pm. The "HomeRx" was an IPC device different from those described in other studies and was used to deliver a constant pressure of 10 mmHg to feet, and sequentially 50 mmHg to ankles, 45 mm Hg to calves and 40 mm Hg to thighs. Sequential pressure was delivered at 2.5 second intervals for a total compression time of 10 seconds followed by 60 seconds of no compression. The control group received compression from Unna's Boots which were changed twice weekly. The method of randomisation was not described. Pre‐trial assessment included ABPI measurement and plethysmographic studies of venous insufficiency measuring venous refilling times and venous filling index. Three patients in the control group experienced adverse reactions; cellulitis (1), allergies (2) and were withdrawn from the study. Ulcer area was measured by tracing the outline onto acetate sheets at the start of the trial and at alternate visits. Leg volumes, amount of wound exudate and pain levels were assessed by direct measurement, subjective grading (light, moderate and heavy) and visual analogue scores respectively.

IPC compared with continuous compression (stockings or bandages)(2 trials, 86 participants)

Two trials (Dolibog 2013; Rowland 2000) compared IPC with compression stockings or bandages.

Dolibog 2013 compared IPC with two separate compression treatments. 22 people were allocated to 12 chamber IPC, from foot to groin. Pressure was set to 60mmHg at the ankle reducing to 40mmHg at the groin. Ventricular filling time was 60 seconds, with a discharge time of 30 seconds. This was delivered daily with each treatment session lasting for 60 minutes. 23 people were allocated an ulcer stocking system delivering pressure of 30‐40mmHg, put on each morning for 10‐12 hours and removed at night. 25 people were allocated to two layer short‐stretch bandaging. Pressure was standardised at 30‐35mmHg or 35‐40mmHg. These were applied each morning for 10‐12 hours and removed at night. All treatments were given for a period of 15 days. In addition, all participants received 2 months of standard drug therapy and gauze dressings saturated in 0.9% sodium chloride changed daily. For the purposes of our analysis all patients receiving either form of continuous compression were grouped together.

Rowland 2000 compared IPC with highly extensible compression bandages worn continuously (Setopress, SSL). Sixteen people with venous leg ulcers were recruited from a wound care outpatient clinic. Randomisation was based on the toss of a coin in this cross‐over trial (data are only included until the point of cross‐over). IPC was delivered using a pump, with pressure set at 50 mmHg, but it was unclear whether a single or multiple cell device was used. Patients were advised to use the IPC device for one hour in the morning and one hour in the evening. People in the control group were treated with high compression, elastomeric bandages. Co‐interventions were the same in each group, ulcers were managed according to a local protocol. Ulcer area was measured by tracing at dressing changes.

Comparison of different IPC regimens (1 RCT, 104 participants)

One trial (Nikolovska 2005) of 104 people compared 'rapid' with 'slow' IPC. 'Rapid' IPC has a sleeve inflation time of 0.5 seconds, a 6 second pressure plateau and 12 seconds deflation. The 'slow' IPC has an inflation time of 60 seconds, a 30 second pressure plateau and 90 seconds deflation time. The IPC device had 7 chambers covering the foot ankle, calf, knee and thigh. Pressure was 45 mmHg at the foot and 30 mmHg at the thigh, for 1 hour daily. Randomisation was by computer generation of a random number sequence, and opening of a sealed opaque, numbered envelope. All patients were treated with a hydrocolloid dressing. Follow up was blinded and was for 6 months. The patients were not treated with compression therapy between IPC sessions. Baseline differences in risk of healing occurred in Nikolovska 2005, where patients in the rapid IPC group tended to have larger ulcers and were of longer duration than those in the slow IPC group (ulcer area mean: 4.47 cm² vs 4.70; ulcer duration mean 4.2 vs 5.4 months). This was the only trial reported using a CONSORT flowchart (Nikolovska 2005).

Risk of bias in included studies

Only one study was overall at low risk of bias (Nikolovska 2005). This study was at low risk of bias in all domains other than blinding of participants and personnel, which was unclear as it was not reported in the paper). A risk of bias table has been completed for each study. Where possible, a quote has been given as evidence for the judgement. The source of the quote is from the primary reference for each study unless otherwise specified.

Allocation

Sequence generation:

All nine studies were described as randomised controlled trials, but only seven reported an adequate method of sequence generation and were at low risk of bias for this domain. Four reported the use of a coin toss method to generate the allocation sequence (Coleridge Smith 1989; Kumar 2002; Nikolovska 2002; Rowland 2000 ), and three studies reported the use of computerised methods to generate the allocation sequence (Alvarez 2012; Dolibog 2013; Nikolovska 2005). One study (McCulloch 1994) randomised patients depending on the last digit of their case note number. This was judged to be at high risk of bias as it is a systematic rather than a truly a random allocation. One study (Schuler 1996) did not state the method of random allocation and was judged to be at high risk of bias for this domain.

Allocation concealment:

Only two studies reported information which confirmed allocation concealment (Dolibog 2013; Nikolovska 2005). Although these studies did not report the envelopes as being opaque, they did report them to be sequentially numbered and sealed and therefore we judged this to be an adequate method of allocation concealment. Remaining seven studies did not provide any information on allocation concealment and hence were judged to be at unclear risk of bias (Alvarez 2012; Coleridge Smith 1989; Kumar 2002; McCulloch 1994: Nikolovska 2002; Rowland 2000; Schuler 1996).

Blinding

Blinding of participants and personnel:

Eight studies were seen as high risk of bias in the domain of blinding participants and personnel (Alvarez 2012; Coleridge Smith 1989; Dolibog 2013; Kumar 2002; McCulloch 1994; Nikolovska 2002; Rowland 2000; Schuler 1996), These studies were judged to be at high risk of bias because no blinding of participants or personnel was described, and due to the nature of the treatment it would be clear if IPC was provided. Nikolovska 2005 was judged to be at unclear risk of bias because no blinding was mentioned, but as two IPC regimens were being compared it is unclear whether blinding was attempted.

Blinding of outcome assessment:

Only Nikolovska 2005 indicated assessment was done by someone unaware of the group (adequate outcome assessor blinding) and was judged to be at low risk of bias. However, one study reported that the outcome assessors were not blinded and hence was judged to be at high risk of bias (Kumar 2002) In six studies it was unclear if the outcome assessors were blinded to the intervention (Alvarez 2012; Coleridge Smith 1989; McCulloch 1994; Nikolovska 2002; Rowland 2000; Schuler 1996). Dolibog 2013 was judged to be at high risk of bias because while the data analysts were blinded to allocation, but the nurse who collected the data was not.

Incomplete outcome data

Kumar 2002 was judged to be at low risk of bias for this domain as the number of drop outs was low (n=6, 15%) and balanced across both groups, Rowland 2000 was judged to be at high risk of bias, as whilst an intention to treat analysis was performed, it had a very high withdrawal rate for a small study (5 out of 16 patients withdrew), Alvarez 2012 was seen as low risk of bias as the number of patients allocated to IPC was reported as 25 in the text and 27 in the tables, this appears to be a transcription error. In Coleridge Smith 1989 the information on drop‐outs was insufficient to make a judgement. The remaining five studies were judged to be at low risk either because missing data was adequately accounted for or there were no patient withdrawals (Dolibog 2013; McCulloch 1994; Nikolovska 2002; Nikolovska 2005; Schuler 1996).

Selective reporting

Reporting bias was judged to be at high risk in Alvarez 2012, as incidence of healing was stated as an outcome measure in the methods but not reported in the results, however this information was supplied in a personal communication with the author. All other studies were judged to be at low risk of reporting bias as all outcome measures stated in the papers were reported. The original protocol was not sought for any of the studies.

Other potential sources of bias

Rowland 2000 was the only study identified to have a high risk of bias from another source, it was a crossover RCT with some patients (n=3 ) crossed over after two months, but others crossed over after three months, though criteria for early crossover is not stated, introducing the possibility for physician to introduce bias.

Effects of interventions

The results are presented according to the specific questions addressed by the review as follows:

(1) Does IPC increase the healing of venous leg ulcers?

Three types of comparison sought to address this question.

(a) trials comparing IPC with no compression
(b) trials comparing IPC plus continuous compression with continuous compression alone
(c) trials comparing IPC with continuous compression provided by bandages or hosiery.

(a) IPC compared with no compression

More people healed in IPC than in dressings alone (25/40, 62.5% compared with 11/40, 27.5%) this was statistically significant. The risk ratio of healing with IPC was 2.27 (95% CI 1.30 to 3.97)(Nikolovska 2002)(Analysis 1.1). This result is at moderate risk of bias since allocation was not secure (coin toss) and outcome assessment was not blinded.

(b) IPC plus continuous compression compared with continuous compression alone

Five trials evaluated the additional effect of IPC over continuous compression (Alvarez 2012; Coleridge Smith 1989; Kumar 2002; McCulloch 1994; Schuler 1996).

Alvarez 2012 reported higher healing rate in the IPC group compared with the IPC group, risk ratio 1.49 (95% CI 1.05 to 2.10)(Analysis 2.1). Coleridge Smith 1989 also reported a significantly higher healing rate in the IPC group (10/21) compared with continuous compression where the healing rate was particularly poor (1/24), risk ratio of healing 11.43 (95% CI 1.59 to 81.99)(Analysis 2.1). It is interesting to note that there was imbalance at baseline in the size of ulcers, the median ulcer area for the group allocated to receive IPC plus continuous compression was much larger than the control group. Kumar 2002 reported no significant difference in proportion of ulcers healed : 20/23 (87%) ulcers healed with IPC, while in the bandaging group 23/25 (92%). The IPC group, in the trial by Kumar, healed more quickly, in an average (mean) of 53.7 days, compared with the 73.7 days to healing in the bandaging group, but these were not formally compared. The rate of reduction in ulcer area was higher in the IPC group (0.14 cm² per day) than in the bandaging group ‐ 0.05 cm² per day. McCulloch 1994 reported no difference in the proportion of ulcers healed in the IPC and the control group; risk ratio (RR) for healing 1.24 (95% CI 0.89 to 1.75)(Analysis 2.1). Schuler 1996 also reported no difference in the proportion of ulcers healed in the IPC and the control group; risk ratio (RR) for healing 1.19 (95% CI 0.80 to 1.77)(Analysis 2.1).

There was clinical heterogeneity in the type of IPC and substantial statistical heterogeneity (I2 = 77%). Undertaking a sensitivity analysis which excluded Coleridge Smith 1989 still resulted in substantial statistical heterogeneity (I2 = 56%). Undertaking a sensitivity analysis which excluded Coleridge Smith 1989 and Kumar 2002 eliminated statistical heterogeneity (I2 = 0%). Excluding Coleridge Smith 1989 and Kumar 2002 and pooling (fixed effect) the remaining three trials showed a statistically significant difference between the two groups (RR 1.31, 95% CI 1.06 to 1.63)(Analysis 3.1).

The duration of IPC evaluated in these trials ranged from 2 to 28 hours per week. Follow‐up period was 3 months in one trial (Coleridge Smith 1989), 4 months in the trial by Kumar 2002, 6 months in two trials (McCulloch 1994; Schuler 1996), and 12 months in one trial (Alvarez 2012). McCulloch 1994 and Kumar 2002 used a single cell IPC device, the remaining three studies used a multiple cell device and described the IPC profile as 'sequential'.

(c) IPC compared with continuous compression (stockings or bandages)

Two trials (Dolibog 2013;Rowland 2000) compared IPC with continuous compression bandages.

Dolibog 2013 recruited 70 participants. No statistically significant difference was detected between IPC and two continuous compression regimes. Risk ratio 1.36 (95% CI 0.50 to 3.70). Rowland 2000 recruited 16 participants and reported no statistically significant difference in healing rates between IPC and continuous compression. No patients were healed in this small study before crossover (Analysis 4.1).

(2) What are the relative effects of different IPC regimens?

Nikolovska 2005 found that rapid IPC healed more ulcers (45/52) than slow IPC (32/52) , RR 1.41, 95%CI 1.11 to 1.79)(Analysis 5.1). The rapid IPC group not only had more healed ulcers, but healing was achieved in a shorter time: median days to complete ulcer healing was 59 with rapid IPC and 100 days with slow IPC (p=0.001; log rank).

(3) Does IPC affect pain or quality of life?

Schuler 1996 reported pain, stating that pain scores, as measured on a visual analogue scale (0‐10) were 3.1 (SD 2.3) with Unna's boot and 2.0 (SD 1.4) with IPC, but that this difference was not statistically significant. The use of different compression systems in this trial, however, confounds the comparison as it is not clear if any difference in pain would be associated with the IPC or the different compression system used. Kumar 2002 stated that comfort was high in both four‐layer and IPC groups (8 to 10 on a 0 to 10 scale which was not validated).

Alvarez 2012 reported pain, measured using a Visual Analogue Scale (0‐10) every week from 0 to 6 weeks. Pain at 6 weeks was 1.5 (SD 1) for IPC and 2.5 (SD 1.3) for compression alone. Pooling Schuler and Alvarez (I² = 0%; Chi² = 0.03, df = 1 (P = 0.87). The difference in pain scores of 1.03 is statistically significant (‐1.03 [95% CI ‐1.56, ‐0.49]) but the clinical impact of a 1 point change in pain scale (VAS 0‐10) is modest Analysis 2.2.

Study Week

0

1

2

3

4

5

6

IPC

Mean and SD

7.8 ± 2.2

5.2 ± 3.1

4.5 ± 2

2.3 ± 2.4

1.9 ± 1.5

2 ± 1

1.5 ± 1

Control

Mean and SD

6.9 ± 2.5

7.1 ± 3.3

6 ± 2.5

4.4 ± 3.1

2 ± 1.3

2.2 ± 1.4

2.5 ± 1.3

Discussion

With a total of nine trials there are 489 participants who contributed data to this review. The studies are on the whole small, and differ in the type (single‐cell compared with sequential: below knee or above knee) and duration of IPC delivered, which does not allow the results to be compared easily. Although one moderately sized trial (Nikolovska 2002) reported that IPC increased healing compared with a dressing alone, this finding is not applicable to the majority of modern settings where compression bandaging is widely used and effective. Were the study to be repeated with people who were intolerant to compression bandaging, then this might provide useful information on a treatment for people who fail to reap the benefit of compression bandages and remain a challenge to practitioners. One small trial reported more ulcers healed with 'fast' IPC therapy when compared with 'slow' IPC therapy. Further trials are needed. The small number of trials means that in‐depth analysis of heterogeneity cannot be explored.

Seven studies described here were undertaken in the community / home care setting and it is unclear whether the place of treatment might have an impact on effectiveness. The other two were undertaken in 'outpatient and inpatient settings at the Clinic of Dermatology, in Skopje in Macedonia.

None of the studies offers a rationale for the chosen dose or duration of IPC although each paper describes potential benefits of the specific method of IPC chosen.

Four trials assessed the impact of adding IPC to a continuous compression regimen ‐ one added IPC to graduated compression stockings (Coleridge Smith 1989), one to Unna's Boot (McCulloch 1994), and two studies compared IPCto four‐layer compression (Alvarez 2012; Kumar 2002). At three months, Coleridge Smith 1989 found a statistically significant difference in the percentage of ulcers healed, whilst at four and six months respectively, Kumar 2002 and McCulloch 1994 found no statistically significant difference in the percentage of ulcers healed. At 12 months Alvarez 2012 found a statistically significant difference in the percentage of ulcers healed. It is not clear whether the differences in outcomes between studies is due to the length of follow‐up, the type of IPC used, the compression therapy used, or chance. Coleridge Smith 1989 reported very poor outcomes in the control group. Kumar suggested that as compression bandaging alone achieves high healing rates there is little benefit to be gained from adding IPC. The study conducted by Schuler 1996 does not allow the independent effect of IPC to be reliably estimated due to the differences in the methods of compression used in the two groups.

Meta‐analysis of the three trials (Alvarez 2012; McCulloch 1994; Schuler 1996) assessing the impact of adding IPC to a continuous compression regimen which minimised statistical heterogeneity shows a statistically significant difference in the number of patients healed (RR 1.41 95% CI 1.06 to 1.63). It should be noted that two of these three studies were rated as being at high risk of bias for random sequence generation.

Meta‐analysis of the two trials reporting pain scores (Alvarez 2012; Schuler 1996) evaluating the impact of adding IPC to a continuous compression regimen on pain shows a small but statistically significant difference in the average pain score, lower pain being associated with IPC (‐1.03, 95% CI ‐1.56, ‐0.49).

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Figuras y tablas -
Figure 1

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Comparison 1 IPC plus dressings versus dressings alone, Outcome 1 Number healed in trial.
Figuras y tablas -
Analysis 1.1

Comparison 1 IPC plus dressings versus dressings alone, Outcome 1 Number healed in trial.

Comparison 2 IPC plus compression versus compression alone, Outcome 1 Number healed in trial.
Figuras y tablas -
Analysis 2.1

Comparison 2 IPC plus compression versus compression alone, Outcome 1 Number healed in trial.

Comparison 2 IPC plus compression versus compression alone, Outcome 2 Pain at 6 weeks.
Figuras y tablas -
Analysis 2.2

Comparison 2 IPC plus compression versus compression alone, Outcome 2 Pain at 6 weeks.

Comparison 3 IPC plus compression versus compression alone ‐ sensitivity analysis, Outcome 1 Number healed in trial.
Figuras y tablas -
Analysis 3.1

Comparison 3 IPC plus compression versus compression alone ‐ sensitivity analysis, Outcome 1 Number healed in trial.

Comparison 3 IPC plus compression versus compression alone ‐ sensitivity analysis, Outcome 2 Rate of healing.
Figuras y tablas -
Analysis 3.2

Comparison 3 IPC plus compression versus compression alone ‐ sensitivity analysis, Outcome 2 Rate of healing.

Comparison 4 IPC versus compression (bandage or stocking), Outcome 1 Number healed in trial.
Figuras y tablas -
Analysis 4.1

Comparison 4 IPC versus compression (bandage or stocking), Outcome 1 Number healed in trial.

Comparison 5 Fast IPC versus Slow IPC, Outcome 1 Number healed in trial.
Figuras y tablas -
Analysis 5.1

Comparison 5 Fast IPC versus Slow IPC, Outcome 1 Number healed in trial.

Comparison 1. IPC plus dressings versus dressings alone

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Number healed in trial Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

Figuras y tablas -
Comparison 1. IPC plus dressings versus dressings alone
Comparison 2. IPC plus compression versus compression alone

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Number healed in trial Show forest plot

5

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected

2 Pain at 6 weeks Show forest plot

2

105

Mean Difference (IV, Fixed, 95% CI)

‐1.03 [‐1.56, ‐0.49]

Figuras y tablas -
Comparison 2. IPC plus compression versus compression alone
Comparison 3. IPC plus compression versus compression alone ‐ sensitivity analysis

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Number healed in trial Show forest plot

3

127

Risk Ratio (M‐H, Fixed, 95% CI)

1.31 [1.06, 1.63]

2 Rate of healing Show forest plot

1

52

Mean Difference (IV, Fixed, 95% CI)

0.90 [‐4.41, 6.21]

Figuras y tablas -
Comparison 3. IPC plus compression versus compression alone ‐ sensitivity analysis
Comparison 4. IPC versus compression (bandage or stocking)

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Number healed in trial Show forest plot

2

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected

Figuras y tablas -
Comparison 4. IPC versus compression (bandage or stocking)
Comparison 5. Fast IPC versus Slow IPC

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Number healed in trial Show forest plot

1

104

Risk Ratio (M‐H, Fixed, 95% CI)

1.41 [1.11, 1.79]

Figuras y tablas -
Comparison 5. Fast IPC versus Slow IPC