Content » Vol 51, Issue 1

Review article

PREDICTORS OF DROPOUT IN INTERDISCIPLINARY CHRONIC PAIN MANAGEMENT PROGRAMMES: A SYSTEMATIC REVIEW

Janke Oosterhaven, PT, MSc1,2, Harriet Wittink, PT, PhD1, Jurgen Mollema, MSc1, Cas Kruitwagen, MSc1,3 and Walter Devillé, MD, PhD2–4

From the 1Research Group Lifestyle and Health, University of Applied Sciences Utrecht, Utrecht, 2Amsterdam Institute for Social Science Research, University of Amsterdam, Amsterdam, 3Julius Centre for Health Sciences and Primary Care, University Medical Centre Utrecht, and 4National Knowledge and Advisory Centre on Migrants, Refugees and Health (Pharos), Utrecht, The Netherlands

Abstract

Objective: Systematic review to identify predictors for dropout during interdisciplinary pain management programmes.

Data sources: PubMed, PsycINFO, CINAHL, Embase, and SPORTDiscus were searched from inception to 22 June 2017.

Study selection: Screening, data-extraction and quality assessment was carried out independently by 2 researchers.

Data synthesis: Eight studies with low methodological quality were included in this review. Out of 63 potential predictors identified in univariate analyses, significant results were found for 18 predictors of dropout in multiple logistic regression analyses in 4 domains, as described by Meichenbaum & Turk: (i) sociodemographic domain (2); (ii) patient domain (8); (iii) disease domain (6); and (iv) treatment domain (2).

Conclusion: This systematic review presents an overview of predictors of dropout. The literature with regard to the prediction of dropout has focused mainly on patient characteristics and is still in the stage of model development. Future research should focus on therapist/therapy-related predictors and the interaction between these predictors. This review suggests future research on this topic, in order to generate better outcomes in interdisciplinary pain management programmes.

Key words: dropout; chronic musculoskeletal pain; interdisciplinary pain management programmes; systematic review.

Accepted Oct 2, 2018; Epub ahead of print Dec 4, 2018

J Rehabil Med 2019; 51: 00–00

Correspondence address: Janke Oosterhaven, Research Group Lifestyle and Health, University of Applied Sciences Utrecht, Heidelberglaan 7, NL-3584 CS Utrecht, The Netherlands. E-mail: janke.oosterhaven@hu.nl

Lay Abstract

Dropout from interdisciplinary pain management programmes is reported to be high. This may result in poor treatment outcomes. Therefore, paying attention to the prevention of dropout is important. This systematic review presents an overview of predictors of dropout (mainly patient characteristics) and makes suggestions for future research on this topic.

Introduction

Dropout from chronic musculoskeletal pain management programmes is associated with poor treatment outcomes (1–3). Despite this, dropout continues to be a neglected topic even after Turk & Rudy brought relapse and non-compliance to attention as a significant problem in chronic pain management (4). This is remarkable, since the assumption is that, because of the complexity of chronic pain management programmes and the focus on self-care behaviour, dropout is inevitable (4–6).

To date, the literature on dropout in interdisciplinary pain management has largely relied on post-hoc analyses on differences between dropouts and treatment completers in cross-sectional research studies. The importance of dropout is acknowledged considering the dropout rates found in these studies ranging from 5% to 46% (4, 7, 8). However, few longitudinal studies have been performed and, as yet, no systematic review has summarized predictors for dropout. More insight in these predictors is needed to develop strategies to prevent premature treatment dropout. Although various definitions of dropout exist, we used the following definition: “patients with chronic pain, who were referred to a chronic pain management programme, who initiated (participated in the baseline assessments), but discontinued prior to completion of the entire programme” (9).

Since therapy for patients with chronic pain is partly similar to that of patients with mental disorders, we can look at mental health research that has been done on dropout in psychotherapy/cognitive therapy. This literature demonstrated dropout incidences around 20% (19.7–21.9%) and valuable insights with regard to predictors of dropout have been gained (10–13). These predictors can be categorized in 5 domains of barriers to retention in therapy, as suggested by Meichenbaum & Turk (14): (i) sociodemographic (dropouts were younger and less educated (10, 13)); (ii) patient (dropouts were less motivated for treatment, preferred alternative treatments and were diagnosed with a specific disorder such as depression or substance use disorder) (10–13); (iii) disease; (iv) treatment (no pre-determined treatment time limit and outpatient treatments were associated with dropout); and (v) healthcare system/system domain (higher dropout rates were found when the treatment was applied by trainees instead of licensed therapists (10)).

The mental health literature demonstrated associations between dropout and negative treatment outcomes. An important reported reason for dropout was the patient’s low motivation. Since research studies select the most motivated patients, selection bias may affect the generalizability of these studies and may lead to overestimation of treatment effects (10–13). It is not clear if these findings may be transferred to interdisciplinary pain management programmes. For most patients with chronic pain these programmes are their last hope. Systematic research on predictors of dropout in these programmes may reveal specific knowledge on how to address these predictors. This may prevent a lot of frustration, overutilization of the healthcare system and more patients who complete interdisciplinary pain management programmes, which may result in better treatment outcomes.

Therefore, the following research question was addressed in this systematic review: which factors are predictors of dropout of patients with chronic musculoskeletal pain during interdisciplinary pain management programmes?

METHODS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (15) and the Cochrane Handbook for Systematic Reviews of Interventions (16) were followed in performing the review. A PROSPERO protocol (registration number: CRD42016039689, https://www.crd.york.ac.uk/prospero/ (17) was developed a priori together with a medical information specialist (JM). Amendments made during the research process were registered in the PROSPERO protocol.

Literature search and study selection

A comprehensive search strategy was conducted from inception to 22 June 2017 in the following databases: PubMed, PsycINFO, CINAHL, Embase and SPORTDiscus. To achieve maximum sensitivity we applied medical subject headings and title/abstract text searches (Appendix S11). The PubMed search strategy was translated to the other databases. Studies obtained through screen-ing reference lists were added in the top righthand box of the PRISMA flowchart (Fig. 1). The search strategies were reviewed and conducted by an experienced medical information specialist.


Fig. 1. Flow chart of study selection.

Screening was done independently by 2 reviewers (JO, HW). The first step was to screen articles on titles and abstracts. The second step was assessment of the full-texts of the articles for inclusion (JO, HW). For each article, any discrepancy between the 2 reviewers was resolved through discussion. In the first screening stage (titles plus abstracts), studies were included when both reviewers agreed they were eligible for inclusion, or if there was doubt about whether or not to exclude them. In the second screening stage (full-texts), studies were included when both reviewers felt they met all the inclusion criteria. The following definition of predictor was used in this systematic review: “a predictive factor is a measurement that is associated with response or lack of response to a particular therapy” (18). Due to the complexity of dropout we focused in this systematic review on studies that applied both univariate analyses and multiple logistic regression analyses. These studies are also known as outcome prediction models (19).

Articles were included if: (i) the study population involved patients over 18 years of age, with chronic non-cancer musculoskeletal pain, i.e. chronic pain that is localized in muscles, ligaments, bones, fasciae, bursae or joints. Chronic pain was defined as: “pain that persists longer than 3 months, or pain that extends beyond the expected period of healing” (20); (ii) the study intervention consisted of interdisciplinary pain management programmes. Interdisciplinary was defined as: “clinicians from different specialities working together and communicating with each other on a frequent and scheduled basis about patients to reach a common goal” (21); (iii) the study design was: retrospective- or prospective cohort study or randomized controlled trials (RCTs) with the aim to identify multiple predictors for dropout during treatment and containing both univariate analyses and multiple logistic regression analyses; (iv) the language of the article was limited to: English, Dutch, French or German. Articles were excluded if the study intervention consisted of: (i) dropout during medication trials or (ii) an online intervention.

Methodological quality assessment

The methodological quality of all included studies was assessed independently by 2 reviewers (JO, HW) with the Quality in Prognosis Studies (QUIPS) tool. This tool is appropriate to assess the risk of bias in prognostic studies and considers 6 domains: study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding, and statistical analyses (Table I) (22, 23).

The reviewers familiarized themselves with the QUIPS through
a test session involving 2 excluded studies, before judging the included studies. All ratings were entered into a spreadsheet. Any difference between the 2 reviewers was resolved through discussion and, if needed, a third reviewer was consulted to reach consensus (WD). An overall score of the study quality was based on the recommendations of Hayden and colleagues (23). For each domain the risk of bias was classified as high, moderate or low. Studies were considered of high quality if in all 6 domains a low risk of bias was found and these studies were labelled as an overall low risk of bias study (23).


Table I. Risk of bias due to various factors, and total score study quality (23)

Data-extraction, data-analyses and data-synthesis

Several steps were taken to extract and synthesize the data from the included studies, all steps independently by 2 reviewers (JO, HW), followed through a discussion, if needed a third reviewer was consulted to reach consensus (WD). In step 1, an extraction manual was designed to facilitate the data-collection process.

The following information was extracted from the included studies: (i) general information: authors, journal, publication date, country, language; (ii) research design: retrospective-or prospective cohort study or RCT; (iii) research population; (iv) analytical approach: univariate analyses using a variety of methods (for example χ2 tests, independent t tests and univariate logistic regression analyses) and multiple logistic regression analyses; (v) all possible factors associated with dropout in univariate analyses and multiple logistic regression analyses with statistical significance and strength of the associations, number of studies that examined the associations.

In step 2 the factors were grouped into 5 domains of   Meichenbaum
& Turk (14).

For each domain the presence of associations and the direction of the associations of predictors and dropout was determined in univariate analyses and multiple logistic regression analyses (Tables II, III, and Table S21 and Table S31).


Table II. Results for (potential) predictors for dropout in simple analysis


Table III. Results for predictors for dropout in multiple logistic regression analysis

For data-synthesis in systematic reviews of studies on outcome prediction models there is still no clear methodological procedure for pooling the data. The heterogeneity of the study populations, study interventions, predictors, statistical analyses and statistical reporting and the fact that most predictors were only investigated in one study (24), did not support applying a best-evidence synthesis (25, 26). Therefore, in step 3, only potential predictors from univariate analyses and multiple logistic regression analyses that were judged in at least 2 studies were described in the results. To summarize the results for a predictor that was investigated in more than 1 study the term: (i) “significant” was assigned if ≥ 75% of the studies showed significant results; (ii) “non-significant” was assigned if ≥ 75% of the studies showed non-significant results; (iii) “conflicting results” was assigned if the rule of ≥ 75% studies showing significant or non-significant could not be applied, or if opposite directions of the association were found in studies (e.g., if dropout was associated with higher pain intensity in one study and with lower pain intensity in another study) (26).

RESULTS
Study selection

The initial search identified a total of 1,954 studies. One additional study was added through screening reference lists (Fig. 1). Without the 555 duplicates, 1,400 studies remained for screening on title and abstract. A total of 32 articles were considered for inclusion, but after full-text screening, only 8 studies were selected for the review. The main reason for exclusion was study design, such as cohort studies with only analyses on differences between completers and dropouts at baseline without prospective or retrospective follow-up and without univariate- or multiple logistic regression analyses of factors that might be predictors for dropout. Two studies were excluded due to the absence of an interdisciplinary approach in the intervention or on the grounds that the intervention under study was an online programme.

Study characteristics

The 8 included studies were conducted between 1994 and 2009. Three studies took place in Europe and 5 studies in the USA. Table S11 provides an overview of the studies included in this review. Most studies focused in their main research objective on detecting predictors of dropout in chronic pain management programmes, 3 studies had a prospective cohort design (24, 27, 28), 4 a retrospective cohort design (29–32) and 1 randomized clinical trial (RCT) with a retrospective secondary analysis on dropout (33).

Interventions

Seven studies described outpatient chronic pain management programmes with an interdisciplinary approach (27, 29, 31–33). Three of these studies were outpatient programmes with a focus on return to work, known as functional restoration programmes (24, 27, 29). One study investigated an inpatient programme with an interdisciplinary approach in the UK (30).

Population

Four studies included patients with chronic low back pain only (27, 29, 32, 33), the other 4 studies included patients with chronic pain in various parts of the body (24, 28, 30, 31). One study conducted a thorough investigation of predictors of dropout with multiple logistic regression analyses in a group of 300 (16%) dropouts from a total sample of 1,845 patients (16%) (24). In 6 studies the dropout group varied from 14 (23%) to 685 patients (16%) (24, 33) and in 2 studies it was unclear how many dropouts were investigated (29, 32).

There were no important differences in sex and in age between 5 of the studies. In 2 studies 70% of the population was female (27, 28) and one study contained a veteran population with 26% females (33). Most studies described patients between 40 and 50 years old, apart from 2 studies, which included patients with a mean age of 35 (29) and 57 years (33).

Reported pain duration varied between a mean of 16 months and 19 years (24, 33) and was not reported in 3 studies (24, 29, 32). The mean pain intensity before treatment ranged from 5 to 8 on a 0–10 NRS scale. One study included only patients with chronic pain who reported worst pain levels of 5 and higher (31).

Terms and definitions and incidence of dropout

Different terms were used to describe dropout: (i) withdrawal from treatment; (ii) early discharge; (iii) non-completion and treatment dropout. Only 4 studies described a definition of dropout (24, 29, 32, 33). The incidence of dropout ranged from 10% in a study in the UK to 51% in a study in the USA (29, 31). Two studies did not report the percentage of dropout (25, 30).

Methodological quality

Analysis of the overall agreement with regard to the methodological quality between the reviewers revealed a Cohen’s weighted kappa of 0.70 (confidence interval (CI) 0.52; 0.88) indicating substantial agreement (34). All studies were considered to be of low quality according to Hayden’s recommendations (23) (Table I). This was due to a serious risk of bias in most studies for study attrition, prognostic factor measurement and statistical
analyses and reporting. For most studies the
reporting of the first step in the statistical analyses, the univariate analyses were poor in contrast with the reporting of the results of the multiple logistic regression analyses (Table S21 predictors organized per study and Table S31: predictors grouped in domains).

Results for predictors for dropout in univariate analyses

In total, 63 potential predictors were studied for an association with dropout in univariate analyses in the 4 domains of the Meichenbaum & Turk: (i) sociodemographic domain (19); (ii) patient domain (21); (iii) disease domain (21) and (iv) treatment domain (2) (Table II). Most potential predictors were examined in a single study. Only 10 out of 63 potential predictors were investigated in more than one study.

Sociodemographic domain

Conflicting results were found for sex (24, 27, 30, 32, 33), ethnicity (24, 32), pre-treatment work-status (24, 33) and job demand (24, 27). Seven of the 8 studies included in this review investigated age as a potential predictor for dropout in univariate analyses. Six out of 7 studies showed significant associations for younger age as a predictor for dropout (27–30, 32, 33).

Patient domain

Conflicting results were found for depression as a potential predictor for dropout. The results of the univariate analyses revealed 2 studies with a significant association of depression with dropout. One study indicated that low depression scores were associated with dropout (28) and another study showed that higher scores on depression scales were associated with dropout (24). Two studies found a non-significant association with dropout (24, 33).

Disease domain

For 2 potential predictors in the disease domain conflicting results in the direction of the association with dropout were found in univariate analyses: for disability (27, 31, 33) and pain intensity (27–31) (Table II). Although significant results were identified for severity of disability and mean pain intensity with dropout in univariate analyses, the direction of the association differed. Only 3 studies showed significant results for pain intensity in association with dropout (28–30). The direction of the association differed in these 3 studies, in one study lower pain intensity (28) and in 2 studies a higher pain intensity was found to be significantly associated with dropout (29, 30).

Results for predictors for dropout in multiple logistic regression analyses

In total 48 of 63 potential predictors were studied for an association with dropout in multiple logistic regression analyses. Of these 48 potential predictors, 26 were not retained in any multiple logistic regression analyses, for 4 predictors conflicting results were found and for 18 predictors significant results were identified in: (i) the sociodemographic domain (2); (ii) patient domain (8); (iii) disease domain (6); and (iv) treatment domain (2). Table III presents an overview of the number of predictors retained and not retained by the multiple logistic regression models (35). Most predictors were found in only a single study (24). Only one predictor, severity of disability, was found in 3 studies (24, 28, 33). Only 2 studies reported results for the performance of the multiple logistic regression models. The Hosmer–Lemeshow test demonstrated in both studies p-values above > 0.5 indicating a good fit (28, 30). No multiple logistic regression models were externally validated using independent samples.

Sociodemographic domain

In multiple logistic regression analyses conflicting results were found for age and pre-treatment work status as potential predictors for dropout. Younger age was not retained in 4 models (27, 30, 32, 33) and was retained in 2 multiple logistic regression models as a predictor for dropout (28, 29). Not working pre-treatment was not retained as a predictor for dropout in one study (33) and was retained in another study (24).

Patient domain

In the patient domain 7 potential predictors that were investigated in univariate analyses in association with dropout were not retained in multiple logistic regression analyses (24, 28, 30). For 8 predictors of dropout significant results were identified in multiple logistic regression models: pre-contemplation, action (28), opioid dependency, any cluster B Dx (24), return to work expectation, somatization (29), self-efficacy and walk distance (30).

Disease domain

For pain intensity and disability conflicting results were demonstrated. In one study lower pain intensity (28) was found to be significantly associated with dropout. In 3 other studies higher pain intensity was identified as a potential predictor for dropout (29–31). Only one of these 3 studies showed significant results in association with dropout in multiple logistic regression analyses (29). Two studies demonstrated that more severe self-reported disability was a significant predictor for dropout (24, 33). Another study found that lower pain disability was significantly associated with dropout (28).

DISCUSSION

The aim of this systematic review was to identify predictors of dropout of patients with chronic musculo-skeletal pain during interdisciplinary pain management programmes. Eight studies with potential predictors for dropout were determined. In total 63 potential predictors were identified in univariate analyses in the 4 domains of retention, as described by Meichenbaum & Turk: (i) sociodemographic domain (19); (ii) patient domain (21); (iii) disease domain (21); and (iv) treatment domain (2). Ten potential predictors (age, sex, social status, education, ethnicity, job demand, depression, pre-treatment work status, pain intensity, and severity of disability) were studied in more than one study and multiple regression analyses revealed conflicting results for almost all these potential predictors.

These conflicting findings are in line with findings known from the mental health literature for the follow-ing predictors: younger age and being diagnosed with a depression (10–12, 27–30, 32, 33). Similar reasons were found in the literature for chronic musculoskeletal pain and mental health, for why younger age may predict dropout from treatment: practical implications, such as having a day-time job or having young children, which may be in conflict with an intensive interdisciplinary treat-ment programme (10, 29, 33). It is known that patients with severe depression, anxiety and low motivation are often excluded from studies about mental health. This may also be the case for studies in this review (13).

Furthermore, this systematic review revealed conflicting findings for pain intensity and disability in association with dropout. An intriguing finding was that one study showed the opposite results for the direction of the association of pain intensity and disability with dropout compared with other studies. The patients who dropped out in this study had lower levels of pain intensity and reported less disability than the completers (28). However, with regard to the unusually high dropout rate of 51% compared with 10–23% in the other studies (24, 30–33), the results of this study may be related to unique elements in this study (poor selection for treatment or inexperienced therapists). Therefore, we may conclude that higher levels of pain intensity and higher severity of disability are predictors for dropout that are worthwhile for future research (24, 29, 33).

Since all studies in this review investigated different combinations of potential predictors, it might not be surprising that conflicting results were demonstrated, for instance, for age, depression, pain intensity, disability and other predictors. Another explanation may be the heterogeneity in study populations due to differences in study design and differences in type of chronic pain treatment (28, 31–33)). The 3 types of interventions identified in this review differed in main programme goal (return to work vs pain management), duration of the programme (4–20 weeks) and selection criteria for the study population. Depending on in which country the study was performed (USA or in European coun-tries), differences in the organization of the healthcare system and referral and funding patterns might have also caused differences in study populations.

Study limitations

The results of this systematic review have to be interpreted in the light of some limitations. First of all, the finding that the methodological quality was considered to be low in all 8 studies (24, 27–33). Quality assessment was a difficult process caused by the lack of consistent and clear reporting of study- and dropout characteristics. Due to a serious risk of bias in most studies for: study attrition and statistical reporting and analysis, the internal validity of the evidence found in this systematic review can be considered to be low. This was probably caused by the retrospective design of most cohort studies (24, 27–33).

Because the lack of conceptualization and operationalization of dropout, the methodological quality of the domain of study attrition of the QUIPS was difficult to judge. Only 4 of the 8 studies provided a definition of dropout (24, 29, 32, 33). This may have influenced the internal validity of this review. The huge differences in dropout rates (10–51%) found in this systematic review can partly be explained by the variation in definitions that were used to describe dropout. A high risk of bias was found for the statistical analysis and reporting for 4 out of 8 studies in this review. Judgement of the statistical results in univariate analyses was most often based on the description in the methods sections of the original articles, as the results were not reported. The statistical analyses in these studies followed a more data-driven approach, which often tend to give a too optimistic estimate of performance of the prediction model and may cause overfitting of the data (36). Only 3 studies reported statistical analyses based on a conceptual framework (28, 29, 31).

The results in this review may suggest that dropout is entirely a patient characteristic (see predictors in the sociodemographic, patient and disease domains). How-ever, as demonstrated in the mental health literature, the experience level of therapists and the therapeutic alliance may be important moderators in association with dropout (10, 12, 13). In this literature it has been recommended that it may be more relevant to focus on the interaction between patient-related predictors and therapist/therapy-related predictors and not on a single key predictor alone (10, 12, 13).

Judging the available evidence in this review, strong limitations concerning the external validity may be raised. A factor that will influence the generalizability of the multiple logistic regression models for dropout presented in this review is related to the differences in chronic pain treatments in the studies. The evidence found for predictors in the 3 different interventions first has to be validated outside the context in which it was gathered (19). For example, most predictors associated with dropout were found in a large prospective cohort study within a functional restoration programme in the USA. These findings may not be generalized to other contexts (24). Another factor that may limit the generalizability of the predictors found in the RCT (33) is the fact that RCTs are known to attract very highly motivated patients, which does not reflect patients with low treatment motivation, who are often seen in daily clinical practice.

It is remarkable that none of the multiple logistic re-gression models that were identified in this review were tested in independent samples. Currently, there is little information available on the performance of the multiple logistic regression models. Three parameters can be described to gain more insight in the performance of the logistic regression models: calibration, discrimination and clinical usefulness (37, 38). Only calibration was described with the goodness-of-fit for 2 multiple logistic regression models in 2 studies (28, 30). Due to the abovementioned limitations, the decision was made not to perform a best-evidence synthesis (25, 26), but to systematically generate a broad overview of all potential predictors found in the literature associated with dropout in chronic pain management programmes. The literature with regard to the topic of dropout is still in the stage of model development, this is defined as type 1A studies, according to the TRIPOD statement (37). Since no internal or external validation was considered in the studies included in this systematic review, the results should be interpreted with care.

Strengths

Since Turk & Rudy addressed the topic of dropout in the early1990s, little high-quality research has become available, as shown by this systematic review. The majority of research on dropout in interdisciplinary pain management programmes for chronic musculoskeletal pain has relied on the investigation of differences between dropouts and programme completers at baseline or univariate analyses of one variable associated with dropout. To our knowledge this is the first systematic review on dropout in chronic musculoskeletal pain patients in the rehabilitation setting. A strength of this review is that it provides an overview of predictors of dropout in chronic musculoskeletal pain: 63 potential predictors of dropout in univariate analyses and 48 out of 63 potential predictors were investigated in multiple logistic regression analyses.

Implications and future directions

Dropout is still a neglected topic in daily practice and in research in interdisciplinary pain management. Learning from the abundant amount of research that has been done in the mental health literature recommendations for future directions can be specified. One direction could be to perform exit interviews with dropouts to investigate reasons for dropout. Based on these reasons for dropout, strategies may be developed to prevent dropout in future (39). Another direction is to address factors that are predictive for dropout in clinical assessment and to make a distinction whether predictors are modifiable or non-modifiable. For example, younger age is a predictor for dropout in chronic pain management programmes. This predictor is non-modifiable and may be addressed by modifying the treatment schedule.

Other strategies that have been developed in the mental health literature that may also be applied in interdisciplinary pain management programmes are: to focus on the assessment of patient’s expectations and preferences and motivation before treatment onset and during the treatment programme, to provide education to address patient’s expectations, preferences and motivation, to monitor the work-alliance between patient and therapist and create the possibility for patients to give feedback to the therapist. Since dropout patients are all different, ideally, prediction of dropout would use prospective cohort designs including session by session evaluation of progress and define no fixed length of the programme (10–13, 39). This systematic review will challenge clinicians and researchers to pay more attention to dropout and collaborate in developing strategies to overcome dropout in order to generate better outcomes in chronic pain management programmes.

Conclusion

This systematic review presents an overview of predictors of dropout. The literature with regard to the prediction of dropout has focused mainly on patient characteristics and is still in the stage of model development. Future research might also focus on therapist/therapy-related predictors and the interaction between these predictors. This review suggests future research on this topic, in order to generate better outcomes in interdisciplinary pain management programmes.

ACKNOWLEDGEMENTS

The authors thank Josca Snoei and Liedeke van Schoot at the University of Applied Sciences Utrecht for their help with the initial construction of this systematic review. The authors also thank Rob Scholten of the Dutch Cochrane Centre for his advice on the best evidence synthesis of this review.

Funding sources: This study is supported by grants of the Dutch government- NWO- 023.004.121 and SIA Raak 2012-14-12P.

Systematic review registration number: CRD42016039689.

The authors have no conflicts of interest to declare.

REFERENCES
  1. Andrews NE, Strong J, Meredith PJ. Activity pacing, avoidance, endurance, and associations with patient functioning in chronic pain: a systematic review and meta-analysis. Arch Phys Med Rehabil 2012; 93: 2109–2121.
    View article    Google Scholar
  2. De Rooij A, Roorda LD, Otten RHJ, van der Leeden M, Dekker J, Steultjens MPM. Predictors of multidisciplinary treatment outcome in fibromyalgia:a systematic review. Disabil Rehabil 2012; 35: 1–13.
    View article    Google Scholar
  3. Van der Hulst M, Vollenbroek-Hutten MMR, IJzerman MJ. A systematic review of sociodemographic, physical, and psychological predictors of multidisciplinary rehabilitation- or, back school treatment outcome in patients with chronic low back pain. Spine 2005; 30: 813–825.
    View article    Google Scholar
  4. Turk DC, Rudy TE. Neglected factors in chronic pain treatment outcome studies-referral patterns, failure to enter treatment, and attrition. Pain 1990; 43: 7–25.
    View article    Google Scholar
  5. Turk DC, Rudy E. Neglected topics in the treatment of chronic pain patients - noncompliance, and adherence enhancement. Pain 1991; 44: 5–28.
    View article    Google Scholar
  6. Horne R, Weinman J, Barber N, Elliott R. Concordance, adherence and compliance in medicine taking: Report for the National Co-ordinating Centre for NHS Service Delivery and Organisation R & D (NCCSDO). London: NCCSDO; 2005.
    View article    Google Scholar
  7. Waterschoot FPC, Dijkstra PU, Geertzen JHB, Reneman MF. Course of disability reduction during a pain rehabilitation program. Int J Rehabil Res 2015; 38: 34–39.
    View article    Google Scholar
  8. McGeary DD, Mayer TG, Gatchel RJ. High pain ratings predict treatment failure in Chronic Occupational Musculoskeletal Disorders. J Bone Joint Surg Am 2006; 88: 317–325.
    View article    Google Scholar
  9. Rainville J, Ahern DK, Phalen. A beliefs about pain and impairment in a functionally oriented treatment program for chronic low back pain. Rainville 1993. Clin J Pain 1993; 9: 196–201.
    View article    Google Scholar
  10. Swift JK, Greenberg RP. Premature discontinuation in adult psychotherapy: a meta-analysis. J Consult Clin Psychol 2012; 80: 547–559.
    View article    Google Scholar
  11. Swift JK, Greenberg RP, Tompkins KA, Parkin SR. Treatment refusal and premature termination in psychotherapy, pharmacotherapy, and their combination: a meta-analysis of head-to-head comparisons. Psychotherapy 2017; 54: 47–57.
    View article    Google Scholar
  12. Fernandez E, Salem D, Swift JK, Ramtahal N, Fernandez E, Salem D, et al. Meta-analysis of dropout from cognitive behavioral therapy: magnitude, timing, and moderators meta-analysis of dropout from cognitive behavioral therapy . J Consult Clin Psychol 2015; 83: 1108.
    View article    Google Scholar
  13. Taylor S, Abramowitz JS, McKay D. Non-adherence and non-response in the treatment of anxiety disorders. J Anxiety Disord 2012; 26: 583–589.
    View article    Google Scholar
  14. Meichenbaum D, Turk DC. Facilitating treatment adherence: a practitioner’s guidebook. New York, NY: Plenum Press; 1987, p. 310.
    View article    Google Scholar
  15. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015; 1–9.
    View article    Google Scholar
  16. Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions version 5.1.0. 2013 [cited 2015 Sep 10] Available from: http://handbook-5-1.cochrane.org/.
    View article    Google Scholar
  17. Oosterhaven J., Wittink H., Mollema, J. Devillé, W. Predictors of dropout of patients with chronic musculoskeletal pain in multidisciplinary pain management programs: a systematic review. PROSPERO 2016 CRD42016039689 [cited 2016 May 27] Available from: http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42016039689.
    View article    Google Scholar
  18. Clark GM. Do we really need prognostic factors for breast cancer?. Breast cancer research and treatment. Breast Cancer Res Tr 1994; 30: 117–126.
    View article    Google Scholar
  19. Van den Berg T, Heymans MW, Leone SS, Vergouw D, Hayden JA, Verhagen AP, et al. Overview of data-synthesis in systematic reviews of studies on outcome prediction models. BMC Med Res Methodol 2013; 13: 42.
    View article    Google Scholar
  20. Merskey H, Bogduk N. Classification of chronic pain, IASP Task Force on Taxonomy. Seattle, WA: International Association for the Study of Pain Press. 1994 [cited 2015 Nov 1] Available from: http://www.iasp-pain.org/PublicationsNews/Content.aspx?ItemNumber=1673.
    View article    Google Scholar
  21. Gatchel RJ, McGeary DD, McGeary CA, Lippe B. Interdisciplinary chronic pain management: past, present, and future. Am Psychol 2014; 69: 119.
    View article    Google Scholar
  22. Hayden JA, Côté P, Bombardier C. Evaluation of the quality of prognosis studies in systematic reviews. Ann Intern Med 2006; 144: 427–437.
    View article    Google Scholar
  23. Hayden JA, van der Windt DA, Cartwright JL, Côté P, Bombardier C. Assessing bias in studies of prognostic factors. Ann Intern Med 2013; 158: 280–286.
    View article    Google Scholar
  24. Howard KJ, Mayer TG, Theodore BR, Gatchel RJ. Patients with chronic disabling occupational musculoskeletal disorder failing to complete functional restoration: analysis of treatment-resistant personality characteristics. Arch Phys Med Rehabil 2009; 90: 778–785.
    View article    Google Scholar
  25. Hayden JA, Tougas ME, Riley R, Iles R, Pincus T. Individual recovery expectations and prognosis of outcomes in non-specific low back pain: prognostic factor exemplar review. The Cochrane Library 2014 [cited 2014 Sep 29] Available from: http:// www.cochrane.org/CD011284/BACK_individual-recoveryexpectations-and-prognosis-of-outcomes-in-non-specificlow-back-pain-prognostic-factor-exemplar-review.
    View article    Google Scholar
  26. Zwikker HE, Van Den Bemt BJ, Vriezekolk JE, Van Den Ende CH, Van Dulmen S, Van Den Bemt B. Psychosocial predictors of non-adherence to chronic medication: systematic review of longitudinal studies. Patient Prefer Adherence 2014; 8: 519–563.
    View article    Google Scholar
  27. Bendix AF, Bendix T, Hæstrup C. Can it be predicted which patients with chronic low back pain should be offered tertiary rehabilitation in a functional restoration program? A search for demographic, socioeconomic, and physical predictors. Spine 1998; 23: 1775–1784.
    View article    Google Scholar
  28. Biller N, Arnstein P, Caudill MA, Federman CW, Guberman C. Predicting completion of a cognitive-behavioral pain management program by initial measures of a chronic pain patient’s readiness for change. Clin J Pain 2000; 16: 352–359.
    View article    Google Scholar
  29. Carosella AM, Lackner JM, Feuerstein M. Factors associated with early discharge from a mulitidisciplinairy work rehabilitation program for chronic tow back pain. Pain 1994; 57: 69–76.
    View article    Google Scholar
  30. Coughlan GM, Ridout KL, Williams ADC, Richardson PH. A from a pain management programme. Coughlan. Br J Clin Psychol 1995; 34: 471–479.
    View article    Google Scholar
  31. Kvaal SA, Lofland KR, Nigro C. Clinical correlates of pain variability in chronic pain. J Back Musculoskelet Rehabil 1999;13: 87–92.
    View article    Google Scholar
  32. Sloots M, Scheppers EF, van de Weg FB, Dekker JH, Bartels EA, Geertzen JH, et al. Higher dropout rate in non-native patients than in native patients in rehabilitation in The Netherlands. Int J Rehabil Res 2009; 32: 232–237.
    View article    Google Scholar
  33. Richmond RL, Carmody TP. Dropout from treatment for chronic low-back pain. Prof Psychol Res Pract 1999; 30: 51.
    View article    Google Scholar
  34. Landis JR, Koch GG. An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics 1977; 363–374.
    View article    Google Scholar
  35. Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific low back pain? A systematic review. Man Ther 2008; 13: 12–28.
    View article    Google Scholar
  36. Moons KGM, de Groot JAH, Bouwmeester W, Vergouwe Y, Mallett S, Altman DG, et al. Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies: The CHARMS Checklist. PLoS Med 2014; 11.
    View article    Google Scholar
  37. Collins, G., Reitsma H, Altman D. Reporting Guideline for Prediction Model Studies : TRIPOD Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis. Ann Intern Med 2015; 162: 55–63.
    View article    Google Scholar
  38. Barrett MS, Chua WJ, Crits-Christoph P, Gibbons MB, Thompson D. Early withdrawal from mental health treatment: implications for psychotherapy practice. Psycho
    View article    Google Scholar
  39. therapy 2008; 45: 247–267.
    View article    Google Scholar
Supplementary content
Appendix S1
Table SI
Table SII
Table SIII

Comments

Do you want to comment on this paper? The comments will show up here and if appropriate the comments will also separately be forwarded to the authors. You need to login/create an account to comment on articles. Click here to login/create an account.
Advertisement