The handling of missing data in trial-based economic evaluations: should data be multiply imputed prior to longitudinal linear mixed-model analyses?

INTRODUCTION: For the analysis of clinical effects, multiple imputation (MI) of missing data were shown to be unnecessary when using longitudinal linear mixed-models (LLM). It remains unclear whether this also applies to trial-based economic evaluations. Therefore, this study aimed to assess whether MI is required prior to LLM when analyzing longitudinal cost and effect data. METHODS: Two-thousand complete datasets were simulated containing five time points. Incomplete datasets were generated with 10, 25, and 50% missing data in follow-up costs and effects, assuming a Missing At Random (MAR) mechanism. Six different strategies were compared using empirical bias (EB), root-mean-squared error (RMSE), and coverage rate (CR). These strategies were: LLM alone (LLM) and MI with LLM (MI-LLM), and, as reference strategies, mean imputation with LLM (M-LLM), seemingly unrelated regression alone (SUR-CCA), MI with SUR (MI-SUR), and mean imputation with SUR (M-SUR). RESULTS: For costs and effects, LLM, MI-LLM, and MI-SUR performed better than M-LLM, SUR-CCA, and M-SUR, with smaller EBs and RMSEs as well as CRs closers to nominal levels. However, even though LLM, MI-LLM and MI-SUR performed equally well for effects, MI-LLM and MI-SUR were found to perform better than LLM for costs at 10 and 25% missing data. At 50% missing data, all strategies resulted in relatively high EBs and RMSEs for costs. CONCLUSION: LLM should be combined with MI when analyzing trial-based economic evaluation data. MI-SUR is more efficient and can also be used, but then an average intervention effect over time cannot be estimated.

Comparing the EQ-5D-5L crosswalks and value sets for England, the Netherlands and Spain: Exploring their impact on cost-utility results.

This study compares the five-level EuroQol five-dimension questionnaire (EQ-5D-5L) crosswalks and the 5L value sets for England, the Netherlands, and Spain and explores the implication of using one or the other for the results of cost-utility analyses. Data from two randomized controlled trials in depression and diabetes were used. Utility value distributions were compared, and mean differences in utility values between the EQ-5D-5L crosswalk and the 5L value set were described by country. Quality-adjusted life years (QALYs) were calculated using the area-under-the-curve method. Incremental cost-effectiveness ratios (ICERs) were calculated, and uncertainty around ICERs was estimated using bootstrapping and graphically shown in cost-effectiveness acceptability curves. For all countries investigated, utility value distributions differed between the EQ-5D-5L crosswalk and 5L value set. In both case studies, mean utility values were lower for the EQ-5D-5L crosswalk compared with the 5L value set in England and Spain, but higher in the Netherlands. However, these differences in utility values did not translate into relevant differences across utility estimation methods in incremental QALYs and the interventions' probability of cost-effectiveness. Thus, our results suggest that EQ-5D-5L crosswalks and 5L value sets can be used interchangeably in patients affected by mild or moderate conditions. Further research is needed to establish whether these findings are generalizable to economic evaluations among severely ill patients.

Diagnostic accuracy of depression questionnaires in adult patients with diabetes: A systematic review and meta-analysis.

BACKGROUND: Comorbid depression is common among patients with diabetes and has severe health consequences, but often remains unrecognized. Several questionnaires are used to screen for depression. A systematic review and meta-analysis regarding the diagnostic accuracy of depression questionnaires in adults with diabetes is unavailable. Our aim was to conduct a systematic review and meta-analysis to evaluate the diagnostic accuracy of depression questionnaires in adults with type 1 or type 2 diabetes. METHODS: PubMed, Embase and PsycINFO were searched from inception to 28 February 2018. Studies were included when the diagnostic accuracy of depression questionnaires was assessed in a diabetes population and the reference standard was a clinical interview. Data extraction was performed by one reviewer and checked by another. Two reviewers independently conducted the quality assessment (QUADAS-2). Diagnostic accuracy was pooled in bivariate random effects models. The main outcome was diagnostic accuracy, expressed as sensitivity and specificity, of depression questionnaires in an adult diabetes population. This study is reported according to PRISMA-DTA and is registered with PROSPERO (CRD42018092950). RESULTS: A total 6,097 peer-reviewed articles were screened. Twenty-one studies (N = 5,703 patients) met the inclusion criteria for the systematic review. Twelve different depression questionnaires were identified, of which the CES-D (n = 6 studies) and PHQ-9 (n = 7 studies) were the most frequently evaluated. Risk of bias was unclear for multiple domains in the majority of studies. In the meta-analyses, five (N = 1,228) studies of the CES-D (≥16), five (N = 1,642) of the PHQ-9 (≥10) and four (N = 822) of the algorithm of the PHQ-9 were included in the pooled analysis. The CES-D (≥16) had a pooled sensitivity of 85.0% (95%CI, 71.3-92.8%) and a specificity of 71.6% (95%CI, 62.5-79.2%); the PHQ-9 (≥10) had a sensitivity of 81.5% (95%CI, 57.1-93.5%) and a specificity of 79.7% (95%CI, 62.1-90.4%). The algorithm for the PHQ-9 had a sensitivity of 60.9% (95%CI, 52.3-90.8%) and a specificity of 64.0% (95%CI, 53.0-93.9%). CONCLUSIONS: This review indicates that the CES-D had the highest sensitivity, whereas the PHQ-9 had the highest specificity, although confidence intervals were wide and overlapping. The algorithm for the PHQ-9 had the lowest sensitivity and specificity. Given the variance in results and suboptimal reporting of studies, further high quality studies are needed to confirm the diagnostic accuracy of these depression questionnaires in patients with diabetes.

Two-year effectiveness of a stepped-care depression prevention intervention and predictors of incident depression in primary care patients with diabetes type 2 and/or coronary heart disease and subthreshold depression: data from the Step-Dep cluster randomised controlled trial.

INTRODUCTION: Major depressive disorders (MDD), diabetes mellitus type 2 (DM2) and coronary heart disease (CHD) are leading contributors to the global burden of disease and often co-occur. OBJECTIVES: To evaluate the 2-year effectiveness of a stepped-care intervention to prevent MDD compared with usual care and to develop a prediction model for incident depression in patients with DM2 and/or CHD with subthreshold depression. METHODS: Data of 236 Dutch primary care patients with DM2/CHD with subthreshold depression (Patient Health Questionnaire 9 (PHQ-9) score ≥6, no current MDD according to the Mini International Neuropsychiatric Interview (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria)) who participated in the Step-Dep trial were used. A PHQ-9 score of ≥10 at minimally one measurement during follow-up (at 3, 6, 9, 12 and 24 months) was used to determine the cumulative incidence of MDD. Potential demographic and psychological predictors were measured at baseline via web-based self-reported questionnaires and evaluated using a multivariable logistic regression model. Model performance was assessed with the Hosmer-Lemeshow test, Nagelkerke's R2 explained variance and area under the receiver operating characteristic curve (AUC). Bootstrapping techniques were used to internally validate our model. RESULTS: 192 patients (81%) were available at 2-year follow-up. The cumulative incidence of MDD was 97/192 (51%). There was no statistically significant overall treatment effect over 24 months of the intervention (OR 1.37; 95% CI 0.52 to 3.55). Baseline levels of anxiety, depression, the presence of >3 chronic diseases and stressful life events predicted the incidence of MDD (AUC 0.80, IQR 0.79-0.80; Nagelkerke's R2 0.34, IQR 0.33-0.36). CONCLUSION: A model with 4 factors predicted depression incidence during 2-year follow-up in patients with DM2/CHD accurately, based on the AUC. The Step-Dep intervention did not influence the incidence of MDD. Future depression prevention programmes should target patients with these 4 predictors present, and aim to reduce both anxiety and depressive symptoms. TRIAL REGISTRATION NUMBER: NTR3715.

Measurement properties of depression questionnaires in patients with diabetes: a systematic review.

PURPOSE: To conduct a systematic review on measurement properties of questionnaires measuring depressive symptoms in adult patients with type 1 or type 2 diabetes. METHODS: A systematic review of the literature in MEDLINE, EMbase and PsycINFO was performed. Full text, original articles, published in any language up to October 2016 were included. Eligibility for inclusion was independently assessed by three reviewers who worked in pairs. Methodological quality of the studies was evaluated by two independent reviewers using the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) checklist. Quality of the questionnaires was rated per measurement property, based on the number and quality of the included studies and the reported results. RESULTS: Of 6286 unique hits, 21 studies met our criteria evaluating nine different questionnaires in multiple settings and languages. The methodological quality of the included studies was variable for the different measurement properties: 9/15 studies scored 'good' or 'excellent' on internal consistency, 2/5 on reliability, 0/1 on content validity, 10/10 on structural validity, 8/11 on hypothesis testing, 1/5 on cross-cultural validity, and 4/9 on criterion validity. For the CES-D, there was strong evidence for good internal consistency, structural validity, and construct validity; moderate evidence for good criterion validity; and limited evidence for good cross-cultural validity. The PHQ-9 and WHO-5 also performed well on several measurement properties. However, the evidence for structural validity of the PHQ-9 was inconclusive. The WHO-5 was less extensively researched and originally not developed to measure depression. CONCLUSION: Currently, the CES-D is best supported for measuring depressive symptoms in diabetes patients.

Effectiveness of a stepped-care intervention to prevent major depression in patients with type 2 diabetes mellitus and/or coronary heart disease and subthreshold depression: A pragmatic cluster randomized controlled trial.

PURPOSE: Given the public health significance of poorly treatable co-morbid major depressive disorders (MDD) among patients with type 2 diabetes mellitus (DM2) and coronary heart disease (CHD), we need to investigate whether strategies to prevent the development of major depression could reduce its burden of disease. We therefore evaluated the effectiveness of a stepped-care program for subthreshold depression in comparison with usual care in patients with DM2 and/or CHD. METHODS: A cluster randomized controlled trial, with 27 primary care centers serving as clusters. A total of 236 DM2 and/or CHD patients with subthreshold depression (nine item Patient Health Questionnaire (PHQ-9) score ≥ 6, no current MDD according to DSM-IV criteria) were allocated to the intervention group (N = 96) or usual care group (n = 140). The stepped-care program was delivered by trained practice nurses during one year and consisted of four sequential treatment steps: watchful waiting, guided self-help, problem solving treatment and referral to the general practitioner. The primary outcome was the 12-month cumulative incidence of MDD as measured with the Mini International Neuropsychiatric Interview (MINI). Secondary outcomes included severity of depression (measured by PHQ-9) at 3, 6, 9 and 12 months. RESULTS: Of 236 patients (mean age, 67,5 (SD 10) years; 54.7% men), 210 (89%) completed the MINI at 12 months. The cumulative incidence of MDD was 9 of 89 (10.1%) participants in the intervention group and 12 of 121 (9.9%) participants in the usual care group. We found no statistically significant overall effect of the intervention (OR = 1.21; 95% confidence interval (0.12 to 12.41)) and there were no statistically significant differences in the course or severity of depressive symptoms between the two groups. CONCLUSIONS: This study suggest that Step-Dep was not more effective in preventing MDD than usual care in a primary care population with DM2 and/or CHD and subthreshold depression.

Process evaluation of a stepped-care program to prevent depression in primary care: patients' and practice nurses' experiences.

BACKGROUND: Depression is common in patients with diabetes type 2 (DM2) and/or coronary heart disease (CHD), with high personal and societal burden and may even be preventable. Recently, a cluster randomized trial of stepped care to prevent depression among patients with DM2 and/or CHD and subthreshold depression in Dutch primary care (Step-Dep) versus usual care showed no effectiveness. This paper presents its process evaluation, exploring in-depth experiences from a patient and practice nurse perspective to further understand the results. METHODS: A qualitative study was conducted. Using a purposive sampling strategy, data were collected through semi-structured interviews with 24 participants (15 patients and nine practice nurses). All interviews were audiotaped and transcribed verbatim. Atlas.ti 5.7.1 software was used for coding and structuring of themes. A thematic analysis of the data was performed. RESULTS: The process evaluation showed, even through a negative trial, that Step-Dep was perceived as valuable by both patients and practice nurses; perceived effectiveness on improving depressive symptoms varied greatly, but most felt that it had been beneficial for patients' well-being. Facilitators were: increased awareness of mental health problems in chronic disease management and improved accessibility and decreased experienced stigma of receiving mental health care. The Patient Health Questionnaire 9 (PHQ-9), used to determine depression severity, functioned as a useful starting point for the conversation on mental health and patients gained more insight into their mental health by regularly filling out the PHQ-9. However, patients and practice nurses did not widely support its use for monitoring depressive symptoms or making treatment decisions. Monitoring mental health was deemed important in chronically ill patients by both patients and practice nurses and was suggested to start at the time of diagnosis of a chronic disease. Appointed barriers were that patients were primarily motivated to participate in scientific research rather than their intrinsic need to improve depressive symptoms. Additionally, various practice nurses preferred offering individually based therapy over pre-determined interventions in a protocolled sequence and somatic practice nurses expressed a lack of competence to recognise and treat mental health problems. CONCLUSION: This study demonstrates both the benefits and unique demands of programs such as Step-Dep. The appointed facilitators and barriers could guide the development of future studies aiming to prevent depression in similar patient groups.

Diagnostic accuracy of the Patient Health Questionnaire-9 for assessment of depression in type II diabetes mellitus and/or coronary heart disease in primary care.

BACKGROUND: Depression is common among type 2 diabetes mellitus (DM2)/coronary heart disease (CHD) patients and is associated with adverse health effects. A promising strategy to reduce burden of disease is to identify patients at risk for depression in order to offer indicated prevention. This study aims to assess the diagnostic accuracy of the Patient Health Questionnaire-9 (PHQ-9) to be used as a tool to identify high risk patients. METHODS: In this cross-sectional study, 586 consecutive DM2/CHD patients aged >18 were recruited through 23 general practices. PHQ-9 outcomes were compared to the Mini International Neuropsychiatric Interview (MINI), which was considered the reference standard. Diagnostic accuracy was evaluated for minor and major depression, comparing both sum- and algorithm based PHQ-9 scores. RESULTS: For minor depression, the optimal cut-off score was 8 (sensitivity 71%, specificity 71% and an AUC of 0.74). For major depression, the optimal cut-off score was 10 resulting in a sensitivity of 84%, a specificity of 82%, and an AUC of 0.88. The positive predictive value of the PHQ-9 algorithm for diagnosing minor and major depression was 25% and 33%, respectively. LIMITATIONS: Two main limitations apply. MINI Interviewers were not blinded for PHQ-9 scores and less than 10% of all invited patients could be included in the analyses. This could have resulted in biased outcomes. CONCLUSIONS: The PHQ-9 sum score performs well in identifying patients at high risk of minor and major depression. However, the PHQ-9 showed suboptimal results for diagnostic purposes. Therefore, it is recommended to combine the use of the PHQ-9 with further diagnostics to identify depression.

Cost-effectiveness of a stepped-care intervention to prevent major depression in patients with type 2 diabetes mellitus and/or coronary heart disease and subthreshold depression: design of a cluster-randomized controlled trial.

BACKGROUND: Co-morbid major depression is a significant problem among patients with type 2 diabetes mellitus and/or coronary heart disease and this negatively impacts quality of life. Subthreshold depression is the most important risk factor for the development of major depression. Given the highly significant association between depression and adverse health outcomes and the limited capacity for depression treatment in primary care, there is an urgent need for interventions that successfully prevent the transition from subthreshold depression into a major depressive disorder. Nurse led stepped-care is a promising way to accomplish this. The aim of this study is to evaluate the cost-effectiveness of a nurse-led indicated stepped-care program to prevent major depression among patients with type 2 diabetes mellitus and/or coronary heart disease in primary care who also have subthreshold depressive symptoms. METHODS/DESIGN: An economic evaluation will be conducted alongside a cluster-randomized controlled trial in approximately thirty general practices in the Netherlands. Randomization takes place at the level of participating practice nurses. We aim to include 236 participants who will either receive a nurse-led indicated stepped-care program for depressive symptoms or care as usual. The stepped-care program consists of four sequential but flexible treatment steps: 1) watchful waiting, 2) guided self-help treatment, 3) problem solving treatment and 4) referral to the general practitioner. The primary clinical outcome measure is the cumulative incidence of major depressive disorder as measured with the Mini International Neuropsychiatric Interview. Secondary outcomes include severity of depressive symptoms, quality of life, anxiety and physical outcomes. Costs will be measured from a societal perspective and include health care utilization, medication and lost productivity costs. Measurements will be performed at baseline and 3, 6, 9 and 12 months. DISCUSSION: The intervention being investigated is expected to prevent new cases of depression among people with type 2 diabetes mellitus and/or coronary heart disease and subthreshold depression, with subsequent beneficial effects on quality of life, clinical outcomes and health care costs. When proven cost-effective, the program provides a viable treatment option in the Dutch primary care system. TRIAL REGISTRATION: Dutch Trial Register NTR3715.