Postoperative braces for degenerative lumbar diseases
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The primary objective is to evaluate the effectiveness of orthosis following lumbar spinal surgery for people with degenerative disease on pain reduction and improvement of functional status. Secondary objectives are to examine the effect of the intervention on overall health or health‐related quality of life and treatment success (from the patient's perspective). We will compare the effect of the intervention to no immobilization.
Background
Description of the condition
Low back pain (LBP) is defined as pain below the last rib and above the inferior gluteal folds either with or without leg pain (Woolf 2003). It is considered chronic when symptoms are present for over three months (Phillips 2013). A source of LBP is only identified in 15% of patients with LBP (Deyo 2001; Wong 2012; Maher 2016). A variety of spinal structures can be identified as sources of pain, including ligaments, facet joints, the vertebral periosteum, the paravertebral musculature and fascia, blood vessels, the anulus fibrosus, and spinal nerve roots. The musculoligamentous injuries and age‐related degenerative processes in the intervertebral disks and facet joints are the most common sources (Deyo 2001; de Schepper 2010).
LBP is the most common reason for missing work in the USA and costs the economy over US$100,000 million per year (Phillips 2013). It is the most burdensome non‐fatal disease and responsible for increasing healthcare costs and productivity loss(Theo Vos 2012). It is the second most common cause for visiting a physician (Deyo 2001) with 2.3% of all doctor visits (Wong 2012). Studies estimate that 77% of people will have LBP and 35% will have sciatica at some point during their life (Heliövaara 1987; Deyo 2001). It is the leading cause of incapacity in adults under 45 years of age and is one of the most frequent cause of early retirement in industrialized countries (Gibson 2007).
Many therapies exist for the treatment of LBP. People are first treated non‐operatively, with either medication, physical therapy, or activity modification, amongst other interventions. If the pain persists or progressive neurological involvement is present, surgery is considered. There is a large number of surgical options available from epidural injections, decompression surgery (instrumented or non‐instrumented), and new technologies, such as disc arthroplasty, interspinous spacer, and intradiscal electrotherapy (Gibson 2005). Fusion is classically indicated in the presence of instability to eliminate painful motion between the two or more vertebrae (Roy‐Camille 1986; Phillips 2013).
Lumbar degenerative disease is the most frequent reason for spinal surgery for which fusion is an option (Deyo 2005; de Schepper 2010). The rate of lumbar fusion is increasing 10 times faster than other orthopaedic procedures, such as total hip or knee replacement (Deyo 2005). Lumbar arthrodesis can be accomplished in a number of ways, for example, through a posterolateral fusion, bone grafting between transverse processes, an interbody technique bone grafting between vertebral bodies, or a combined approach (Fritzel 2002).
Description of the intervention
There is evidence of lumbosacral orthosis being used since biblical times for aesthetic purposes and only since 1530 AD for external lumbar support (Fidler 1983). Its use, nowadays, varies from management of LBP (Ahlgren 1978; Sato 2012) to restricting the movement of an operated spine (Lorenz 1991; Johnsson 1992; Fritzel 2002; Bible 2009).
Many types of orthosis have been created for medical purposes. The most commonly used are rigid, flexible, corset and jacket, lumbar, thoracolumbar, and lumbosacral, and all with the intention to provide more stability to the spine (Ahlgren 1978; Fidler 1983; Lorenz 1991; Johnsson 1992; Fritzel 2002; Bible 2009; Sato 2012).
The first description of orthosis following spinal surgery was reported by Hibbs 1924. The technique used was posterolateral fusion without instrumentation followed by a full‐time external immobilization with rigid braces for approximately six months to one year.
With the advent of spinal implants, such as hooks and pedicle screws, surgeons have been able to provide rigid internal stabilization of spine. The improvement in fusion rates with the new generation of spinal implants has been demonstrated in a prospective study of a one‐level fusion with and without instrumentation for disabling back pain (Lorenz 1991). The author reported superior results to instrumented fusions. In addition, there were no reports of pseudarthrosis among the instrumented group. However, even with this evidence, external immobilization is still being used as a complement to the surgery (Johnsson 1992; Fritzel 2002; Bible 2009).
The controversial question is whether braces reduce movement in the operated spine causing higher rates of consolidation (Johnsson 1992). Some authors believe that pain following a surgery is less severe when an external brace is applied, which may be due to increased stability from the brace, and it also gives the patients a psychological sense of security following the procedure (Connolly 1998; Bible 2009).
Potential complications associated with postoperative spinal orthosis include skin discolouration, pressure sores, injury to the lateral femoral cutaneous nerve, allergic reaction to bracing material, and heat retention causing skin maceration. In addition, the cost of a custom‐moulded orthosis can be high (Connolly 1998; Bible 2009).
How the intervention might work
External postoperative orthoses may improve patient outcomes by allowing less movement to the segment operated and creating more stability to the spine, causing less pain, allowing better healing, and increasing the fusion rates (Fidler 1983; Connolly 1998; Bible 2009).
Why it is important to do this review
Postoperative orthoses are widely prescribed for postoperative purposes. However, there is a lack of consensus regarding the most appropriate type, duration, and indications for use (Johnsson 1992). There is no consistent scientific evidence that suggests use of a thoracolumbosacral orthosis may significantly reduce lumbar intervertebral mobility, particularly in the lower lumbar spine, and improve the patient's outcome (Fidler 1983; Connolly 1998; Bible 2009).
Objectives
The primary objective is to evaluate the effectiveness of orthosis following lumbar spinal surgery for people with degenerative disease on pain reduction and improvement of functional status. Secondary objectives are to examine the effect of the intervention on overall health or health‐related quality of life and treatment success (from the patient's perspective). We will compare the effect of the intervention to no immobilization.
Methods
Criteria for considering studies for this review
Types of studies
We will include randomized controlled trials (RCTs) and quasi‐RCTs, as defined by those studies that use a method of treatment allocation that is not strictly random e.g. date of birth, hospital record number, or alternation. Studies will be eligible regardless of the language or date of publication. We will only consider trials published as full‐text articles or available as a full trial report for inclusion.
Types of participants
We will include adults (greater than 18 years of age) with degenerative lumbar disease and degenerative instability who have undergone lumbar surgery. We will include patients with one or more of the following diagnoses: degenerative disc disease, lumbar disc herniation, lumbar stenosis, degenerative scoliosis, spondylolysis, or spondylolisthesis.
We will not place any restrictions on the type of graft (autograft or allograft) or spinal implants used to achieve fusion. We will exclude studies with patients with documented osteoporoses, infection, fixed imbalance correction surgery, or fusion extended to iliac bone.
Types of interventions
We will include studies that examine postoperative lumbar orthoses, and we will not place any restriction on the type, i.e. whether it is rigid or flexible, lumbar, thoracolumbar, or lumbosacral. We will include studies that employ co‐interventions in addition to orthoses provided we can determine the effect of the intervention, i.e. as long as potential co‐interventions are equivalent in both the intervention and comparison group.
Types of outcome measures
Primary outcomes
-
Back‐specific functional status (Oswestry Disability Index (ODI) (Fairbank 2000; Roland 2000)), Roland‐Morris Questionnaire (RMQ) (Roland 1983; Roland 2000)).
-
Back pain (as measured by Visual Analogue Scale (VAS)).
Secondary outcomes
-
Quality of life (general health questionnaire, Short‐Form 36 (SF‐36) (Garratt 1993)).
-
Use of analgesic medications.
-
Failure and complications in postoperative period (neurological, vascular, infection, hardware failure).
-
Return‐to‐work.
We will extract outcome measures that assess the benefits of treatment at the following time points: short‐term (less than three months after randomization), medium‐term (three to six months after randomization), and long‐term (greater than six months after randomization).
Search methods for identification of studies
One co‐author of this review is an experienced librarian (AY) and she will assist the team with the development of the electronic search strategies.
Electronic searches
We will search the following databases from inception to the current period for relevant trials.
-
The Cochrane Back and Neck Trials Register (CENTRAL or the Cochrane Register of Studies (CRS), or both).
-
The Cochrane Central Register of Controlled Trials (CENTRAL, the Cochrane Library, latest issue).
-
MEDLINE (Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE) (OvidSP, 1946 to present).
-
Embase (OvidSP, 1980 to present).
-
Web of Science (Thomson Reuters, 1900 to present).
-
Latin American and Caribbean Health Sciences Literature (LILACS) (1982 to present).
-
The World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP)
The search strategy will be composed of terms for the intervention (orthosis) and the condition (degenerative disorders) in order to maximize sensitivity. As we will search both subject headings and free text words, we expect that we will identify all relevant studies.
A draft MEDLINE search strategy is in Appendix 1. We will adapt the search strategy as closely as possible to the other databases listed above.
Searching other resources
We will also screen the reference lists of the included studies and all related (systematic) reviews; perform citation tracking; and contact specialists in the field and authors of the included trials for information on unpublished data.
Data collection and analysis
For study selection and data extraction, we will follow a standard protocol as recommended by Cochrane Back and Neck's (CBN) method guideline (Furlan 2015). Two review authors (AOG and ANM) will independently conduct all steps that are potentially prone to bias. This includes the selection of studies, 'Risk of bias' assessments, and data extraction. We will use a standardized data extraction form. We will resolve any discrepancies first by discussion and, when necessary, a third review author (SMR) will act as arbiter. Review authors will not be involved in data extraction and the analyses for studies of which they are a co‐author.
In all cases where the studies provide insufficient data and there is uncertainty, we will attempt to contact the corresponding study author(s) for clarification.
Selection of studies
Two review authors (AOG and ANM) will independently screen the titles and abstracts of all reports identified by the literature searches and will include those studies based upon the study design, types of participants, and type of intervention as defined in the inclusion criteria. We will obtain and assess the full‐text articles of all studies that appear to be potentially relevant. We will list all studies excluded after full‐text assessment and their reasons for exclusion in a 'Characteristics of excluded studies' table. Also we will construct a PRISMA flow diagram to illustrate the study selection process.
Data extraction and management
We will extract the following data from the included studies.
-
Study characteristics (e.g. country where the study was conducted, patient population source or setting, methods of recruitment and randomization, sources of funding, inclusion criteria.
-
Population characteristics (e.g. number of patients, age, gender, duration of LBP, co‐morbidity).
-
Intervention characteristics (e.g. description of the modality: rigid or flexible orthosis, duration of time that the orthosis must be worn, co‐interventions).
-
Outcome measures (e.g. pain, functional status, overall health, adverse events).
We will enter all extracted data into Review Manager 5 (RevMan 5).
Assessment of risk of bias in included studies
Two review authors will independently assess the risk of bias in the included studies using CBN recommendations (Furlan 2015).
We will assess the following sources of bias, as presented in Table 1: selection bias; performance bias; detection bias; attrition bias; reporting bias; and other bias. The criteria for a judgement of 'yes' for the 'Risk of bias' source are in Table 2.
| Bias domain | Source of bias | Possible answers |
| Selection | (1) Was the method of randomisation adequate? | Yes/no/unsure |
| Selection | (2) Was the treatment allocation concealed? | Yes/no/unsure |
| Performance | (3) Was the patient blinded to the intervention? | Yes/no/unsure |
| Performance | (4) Was the care provider blinded to the intervention? | Yes/no/unsure |
| Detection | (5) Was the outcome assessor blinded to the intervention? | Yes/no/unsure |
| Attrition | (6) Was the drop‐out rate described and acceptable? | Yes/no/unsure |
| Attrition | (7) Were all randomized participants analysed in the group to which they were allocated? | Yes/no/unsure |
| Reporting | (8) Are reports of the study free of suggestion of selective outcome reporting? | Yes/no/unsure |
| Selection | (9) Were the groups similar at baseline regarding the most important prognostic indicators? | Yes/no/unsure |
| Performance | (10) Were co‐interventions avoided or similar? | Yes/no/unsure |
| Performance | (11) Was the compliance acceptable in all groups? | Yes/no/unsure |
| Detection | (12) Was the timing of the outcome assessment similar in all groups? | Yes/no/unsure |
| Other | (13) Are other sources of potential bias unlikely? | Yes/no/unsure |
Table extracted from Furlan 2015
| 1 | A random (unpredictable) assignment sequence. Examples of adequate methods are coin toss (for studies with 2 |
| 2 | Assignment generated by an independent person not responsible for determining the eligibility of the patients. |
| 3 | Index and control groups are indistinguishable for the patients or if the success of blinding was tested among |
| 4 | Index and control groups are indistinguishable for the care providers or if the success of blinding was tested |
| 5 | Adequacy of blinding should be assessed for each primary outcome separately. This item should be scored ‐ for patient‐reported outcomes in which the patient is the outcome assessor (e.g., pain, disability): the blinding |
| 6 | The number of participants who were included in the study but did not complete the observation period or |
| 7 | All randomized patients are reported/analyzed in the group they were allocated to by randomization for the |
| 8 | All the results from all prespecified outcomes have been adequately reported in the published report of the |
| 9 | Groups have to be similar at baseline regarding demographic factors, duration and severity of complaints, |
| 10 | If there were no cointerventions or they were similar between the index and control groups. |
| 11 | The reviewer determines if the compliance with the interventions is acceptable, based on the reported |
| 12 | Timing of outcome assessment should be identical for all intervention groups and for all primary outcome |
| 13 | Other types of biases. For example: |
Table extracted from Furlan 2015
We will record the information for each included trial in 'Risk of bias' tables in RevMan 5 (RevMan 5), and will summarize the risk of bias for each included study in a summary 'Risk of bias’ figure and graph.
Measures of treatment effect
We will use RevMan 5 to perform the data analyses (RevMan 5).
In order to be effective, the brace should be worn exactly as prescribed by the surgeon. Therefore, our analyses will be limited to patients who used the braces for the duration prescribed by the healthcare provider. Patients who have not used the braces for the minimum prescribed duration will be excluded from the analyses. We will report the compliance with treatment as the number of patients who used the braces exactly as prescribed by the surgeon divided by the total number of patients randomized.
We will express continuous outcomes as mean differences (MDs) with 95% confidence intervals (CIs) when different trials use the same measurement instrument to measure the same outcome. Otherwise we will use a standardized mean difference (SMD) values when trials measure the same outcome but employ a different measurement instrument to measure the same conceptual construct (e.g. functional status). We will express dichotomous outcomes as a risk ratio (RR) with 95% CIs.
To enhance interpretability of dichotomous outcomes, we will calculate absolute risk differences. For outcomes with statistically significant differences between intervention groups, we will determine the number needed to treat for an additional beneficial outcome (NNTB) or the number needed to treat for an additional harmful outcome (NNTH). To enhance the interpretability of continuous outcomes, we will back‐transform the pooled SMD values of overall pain and disability to an original 0 to 10 mm VAS for pain and to a common metric of 0 to 100 for disability by multiplying the SMD and 95% CIs by a representative pooled standard deviation (SD) at baseline of one of the included trials, which we will base upon similar populations, similar risk of bias, and type of intervention.
Unit of analysis issues
The unit of analysis will be each patient recruited to the included trials.
Dealing with missing data
We will contact the authors of primary studies when data are missing or incomplete, such as numbers of participants, details of dropouts, means, measures of variance (SD or standard error (SE)), and numbers of events. If the needed data are not available, or if data does not appear in a traditional format, we will use the data available to appropriately calculate missing values.
For dichotomous outcomes that measure benefits (e.g. proportion of participants reporting pain relief of 30% or greater), we will calculate the worst‐case analysis using the number of randomized patients as the denominator. For continuous outcomes (e.g. pain, functional status), we will calculate the MD or SMD based on the number of patients analysed at that time point. If the number of patients analysed is not presented for each time point, we will use the number of randomized patients in each group at baseline.
For continuous outcomes with no reported SD, we will calculate SD values from SE values, 95% CIs, or P values, if the included studies present these values. If no measures of variation are reported and we are unable to calculate SD values, we plan to impute SD values from other included trials in the same meta‐analysis, using the median of the other SDs available (Akl 2013). For continuous outcomes that are only presented graphically, we will extract the mean and 95% CIs from the graphs visually, where possible. For dichotomous outcomes, we will use percentages to estimate the number of events or the number of participants assessed for an outcome.
Where data are imputed or calculated (e.g. SDs calculated from SEs, 95% CIs, or P values, or imputed from graphs or from SDs in other trials), we will report this in the 'Characteristics of included studies' tables and we will further analyse these effects in sensitivity analyses.
Assessment of heterogeneity
We will assess heterogeneity by visual inspection of the forest plots, the Q‐test and the I² statistic (Deeks 2011). We will classify heterogeneity using the following I² statistic values.
-
0 to 40%: might not be important.
-
30% to 60%: may represent moderate heterogeneity.
-
50% to 90%: may represent substantial heterogeneity.
-
75% to 100%: considerable heterogeneity.
In cases of considerable heterogeneity (defined as an I² statistic value ≥ 75%), we will explore the data further, and will compare the characteristics of individual studies and will conduct subgroup analyses.
Assessment of reporting biases
If there is more than 10 included studies, we plan to draw funnel plots of primary outcomes to assess potential publication bias.
We will assess the presence of small study bias in the overall meta‐analysis by the random‐effects meta‐analysis of the intervention. In the presence of small study effects, the random‐effects model will give a more beneficial estimate of the intervention than the fixed‐effect model (Sterne 2011).
Pain, functional and complications are typically reported outcomes for LBP. In order to assess selective outcome reporting bias, we will compare the trial protocols of all included studies with the outcomes reported in the individual studies. In the absence of a protocol, we will downgrade the study
Data synthesis
Regardless of whether there are sufficient data available to quantitatively summarize the data, we will assess the overall quality of the evidence for each outcome. To accomplish this, we will use the GRADE approach (GRADEpro), as recommended in theCochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), and adapted in the updated CBN method guidelines (Furlan 2015). The quality of the evidence on a specific outcome is based on the performance of the studies against five factors: study design and risk of bias, inconsistency of results, indirectness (not generalizable), imprecision (few events and few patients), and other factors (e.g. reporting bias).
For RCTs the quality of the evidence starts as high and is reduced by a level for each of the factors not met, as described in Appendix 2.
'Summary of findings' tables
We will use one or more 'Summary of findings' tables in our review (Table 3), which provide the information concerning the quality of evidence, and the magnitude and precision of the effect of the intervention. We will include the outcomes in the 'Summary of findings' tables and relate these to the timing of the measurement. The major comparison will be better results with and without orthesis after surgery. We will include the outcomes of function, pain, quality of life, use of analgesic medications, postoperative failure and complications, and return to work in the 'Summary of findings' tables.
| Postoperative braces versus no postoperative braces for adults with degenerative lumbar diseases | |||
| Patient or population: adults with degenerative lumbar disease Settings: postoperative of degenerative lumbar disease Intervention: with postoperative braces Comparison: without postoperative braces | |||
| Outcomes | Outcome type (continuous/ dichotomous) | Outcome measure | Comments |
| Function | Continuous | Oswestry Disabilty Index (ODI), Roland‐Morris Questionnaire (RMQ) | medium‐term |
| Back pain | Continuous | Visual Analogue Scale (VAS) | medium‐term |
| Quality of life | Continuous | SF‐36 questionnaire | medium‐term |
| Use of analgesic medications | Dichotomous | Yes/no | short‐term |
| Postoperative complications | Dichotomous | Yes/no | medium‐term |
| Return to work | Dichotomous | Yes/no | long‐term |
| GRADE Working Group grades of evidence | |||
| Short‐term: less than 3 months after surgery Medium‐term: 3 to 6 months after surgery Long‐term: greater than 6 months after surgery | |||
We will assess functional status by back‐specific questionnaires, ODI (Fairbank 2000; Roland 2000), and RMQ (Roland 1983; Roland 2000); back pain by a VAS; and quality of life by the SF‐36 questionnaire (Garratt 1993).
Timing of these outcome assessments will be short‐term (less than three months after surgery) for the use of analgesic medication; medium‐term (three to six months after operation) for function, back pain, quality of life and postoperative complications, since the time for the arthrodesis to consolidate is around three months; and long‐term (greater than six months) for return to work, to evaluate the success of the surgery with and without the orthosis.
Subgroup analysis and investigation of heterogeneity
Subgroup analyses are secondary analyses in which the participants are divided into groups according to shared characteristics and outcomes to determine if any significant treatment effect occurs according to that characteristic. If data permit, we will perform the following subgroup analyses.
-
Type of graft (allograft versus autograft).
-
Type of orthosis used postoperatively (rigid or flexible, lumbar, thoracolumbar, or lumbosacral).
-
Surgery submitted: discectomy, fusion, laminectomy, and others.
-
Type of fusion performed (pedicle screws only versus concomitant interbody graft with cage).
Sensitivity analysis
If there is an adequate number of included studies, we will perform a sensitivity analysis to explore the causes of heterogeneity and the robustness of the results. We will conduct the following sensitivity analyses.
-
Risk of bias (selection and attrition bias).
-
Type of orthosis used postoperatively (rigid or flexible, lumbar, thoracolumbar, or lumbosacral).
-
Assumptions surrounding missing data (e.g. those studies where data are imputed versus complete data sets).
-
Publication bias and conflicts of interest (i.e. the granting agency)
-
Type of graft (allograft versus autograft).
-
Surgery submitted: discectomy, fusion, laminectomy, and others.
-
Type of fusion performed (pedicle screws only versus concomitant interbody graft with cage)
| Bias domain | Source of bias | Possible answers |
| Selection | (1) Was the method of randomisation adequate? | Yes/no/unsure |
| Selection | (2) Was the treatment allocation concealed? | Yes/no/unsure |
| Performance | (3) Was the patient blinded to the intervention? | Yes/no/unsure |
| Performance | (4) Was the care provider blinded to the intervention? | Yes/no/unsure |
| Detection | (5) Was the outcome assessor blinded to the intervention? | Yes/no/unsure |
| Attrition | (6) Was the drop‐out rate described and acceptable? | Yes/no/unsure |
| Attrition | (7) Were all randomized participants analysed in the group to which they were allocated? | Yes/no/unsure |
| Reporting | (8) Are reports of the study free of suggestion of selective outcome reporting? | Yes/no/unsure |
| Selection | (9) Were the groups similar at baseline regarding the most important prognostic indicators? | Yes/no/unsure |
| Performance | (10) Were co‐interventions avoided or similar? | Yes/no/unsure |
| Performance | (11) Was the compliance acceptable in all groups? | Yes/no/unsure |
| Detection | (12) Was the timing of the outcome assessment similar in all groups? | Yes/no/unsure |
| Other | (13) Are other sources of potential bias unlikely? | Yes/no/unsure |
| Table extracted from Furlan 2015 | ||
| 1 | A random (unpredictable) assignment sequence. Examples of adequate methods are coin toss (for studies with 2 |
| 2 | Assignment generated by an independent person not responsible for determining the eligibility of the patients. |
| 3 | Index and control groups are indistinguishable for the patients or if the success of blinding was tested among |
| 4 | Index and control groups are indistinguishable for the care providers or if the success of blinding was tested |
| 5 | Adequacy of blinding should be assessed for each primary outcome separately. This item should be scored ‐ for patient‐reported outcomes in which the patient is the outcome assessor (e.g., pain, disability): the blinding |
| 6 | The number of participants who were included in the study but did not complete the observation period or |
| 7 | All randomized patients are reported/analyzed in the group they were allocated to by randomization for the |
| 8 | All the results from all prespecified outcomes have been adequately reported in the published report of the |
| 9 | Groups have to be similar at baseline regarding demographic factors, duration and severity of complaints, |
| 10 | If there were no cointerventions or they were similar between the index and control groups. |
| 11 | The reviewer determines if the compliance with the interventions is acceptable, based on the reported |
| 12 | Timing of outcome assessment should be identical for all intervention groups and for all primary outcome |
| 13 | Other types of biases. For example: |
| Table extracted from Furlan 2015 | |
| Postoperative braces versus no postoperative braces for adults with degenerative lumbar diseases | |||
| Patient or population: adults with degenerative lumbar disease Settings: postoperative of degenerative lumbar disease Intervention: with postoperative braces Comparison: without postoperative braces | |||
| Outcomes | Outcome type (continuous/ dichotomous) | Outcome measure | Comments |
| Function | Continuous | Oswestry Disabilty Index (ODI), Roland‐Morris Questionnaire (RMQ) | medium‐term |
| Back pain | Continuous | Visual Analogue Scale (VAS) | medium‐term |
| Quality of life | Continuous | SF‐36 questionnaire | medium‐term |
| Use of analgesic medications | Dichotomous | Yes/no | short‐term |
| Postoperative complications | Dichotomous | Yes/no | medium‐term |
| Return to work | Dichotomous | Yes/no | long‐term |
| GRADE Working Group grades of evidence | |||
| Short‐term: less than 3 months after surgery Medium‐term: 3 to 6 months after surgery Long‐term: greater than 6 months after surgery | |||
