Interventions to increase attendance for diabetic retinopathy screening: systematic review and meta-analysis

Methodological quality of the review: High confidence

Author: Lawrenson JG, Graham-Rowe E, Lorencatto F, Rice S, Bunce C, Francis JJ, Burr JM, Aluko P, Vale L, Peto T, Presseau J, Ivers NM, Grimshaw JM.

Region: United States (US), Canada, Australia, Germany, United Kingdom (UK) and Netherlands.

Sector: Diabetic retinopathy screening

Sub-sector: interventions, uptake, cost-effectiveness

Equity focus: None specified

Review type: Other review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

The majority of studies assessing the effectiveness of QI interventions to improve diabetes care (including those delivered specifically to improve DRS) involve multicomponent interventions that attempt to change the behaviour of HCPs (e.g. advising patients to attend DRS) or patients (e.g. actually attending) or both. As there is no consistent association between the number of intervention components and their effectiveness, the ‘ideal’ number of components in such programmes is unknown. Furthermore, given the complexity of interventions tested to date, it is not always clear which specific components are the effective elements of these interventions (i.e. the ‘active ingredients’).

Objectives:

Systematically review the evidence from RCTs for the effectiveness and cost-effectiveness of QI interventions that seek to increase attendance for DRS.

Enrich the data set by contacting authors of included studies to obtain information on missing data relating to the content of the intervention and/or context.

Code descriptions of the interventions reported in the included RCTs in terms of the type of QI interventions used and their constituent BCTs.

Explore heterogeneity in effect size using conventional and innovative meta-analytic methods to identify factors (including BCTs) associated with greater effectiveness.

Main findings:

In total, authors included 66 studies in the review. Of the included studies, 50 reported general QI interventions and evaluated the impact of the interventions across a range of outcomes, including DRS uptake. In 16 of the included studies, the primary target of the intervention was to improve attendance for DRS. In addition authors identified nine ongoing trials. 35 studies were paralleled-group patient RCTs and 31 were cluster RCTs in which HCP or the health-care setting was the unit of randomization.

Studies were all conducted in high income countries, including US, UK, Australia, Canada, Germany and Netherlands. However, 17 studies were conducted in disadvantaged populations. Interventions were either specially targeted at improving attendance (16) or were part of a general QI intervention to improve diabetes care (50). Out of the 16 studies, in 12 of the studies the outcome was a dilated fundus examination (DFE). In the 50 studies DRS was usually listed as part of a number of processes of care based on diabetes guideline recommendations. Generally, interventions were multifaceted with several QI components per intervention arm (median 3, range 1–7). For interventions specifically targeting DRS attendance, the most commonly used QI components were ‘patient reminders’ (56% of studies) and ‘patient education’ (75%). For general QI interventions, a greater number and range of strategies were used, including ‘patient education’ (48% of studies), ‘promotion of self-management’ (40%), ‘case management’ (40%), ‘clinician education’ (38%) and ‘team changes’ (36%).

Overall, trials were assessed low risk of bias, however, 33 were attributed high risk of bias. Based on the forest plot observations, authors note that although there were a few data points for the less precise studies, those with greater precision were evenly distributed.

In terms of economic outcomes, five studies reported full economic evaluation: three of these were cost-effectiveness analyses and two were cost–consequence analyses. Nine studies reported partial economic evaluations: five were resource utilisation studies and four were cost outcome descriptions. The risk of bias of these studies were similar to the risk of bias of the main body of the included studies.

Overall, interventions increased DRS attendance by 12% [risk difference (RD) 0.12, 95% confidence interval (CI) 0.10 to 0.14] compared with usual care, with substantial heterogeneity in effect size. Both DRS-targeted and general QI interventions were effective, particularly when baseline attendance levels were low. All commonly used QI components and BCTs were associated with significant improvements, particularly in those with poor attendance. Higher effect estimates were observed in subgroup analyses for the BCTs of ‘goal setting (outcome, i.e. consequences)’ (RD 0.26, 95% CI 0.16 to 0.36) and ‘feedback on outcomes (consequences) of behaviour’ (RD 0.22, 95% CI 0.15 to 0.29) in interventions targeting patients and of ‘restructuring the social environment’ (RD 0.19, 95% CI 0.12 to 0.26) and ‘credible source’ (RD 0.16,95% CI 0.08 to 0.24) in interventions targeting HCPs.

Phase 2 – 3457 studies were screened, of which 65 non-randomised studies were included in the review. The following theoretical domains were likely to influence attendance: ‘environmental context and resources’, ‘social influences’, ‘knowledge’, ‘memory, attention and decision processes’, ‘beliefs about consequences’ and ‘emotions’.

Methodology: 

Authors considered RCTs, both individually randomized and cluster RCTs, conducted in a primary or a secondary care setting. To investigate the cost-effectiveness, authors included full economic evaluations (cost-effectiveness analysis, cost-utility analyses and cost-benefit analyses), cost-analyses and comparative resource utilization studies conducted alongside or as part of an included RCT. Authors included participants with type 1 and type 2 diabetes mellitus who were eligible for DRS. Interventions consisted of RCTs that used any planned strategy or combination of strategies to improve attendance for DRS targeted at individuals with diabetes, HPCs or the health-care system. Interventions included those specially targeting DRS as well as those that were part of a general WI interventions for diabetes care. Comparator interventions were as specific in the included studies.

Primary outcome measure was one or more visits for DRS within a 2-year period following randomization. This could be based on self-reports, medical insurance claims databases or health record audits. Secondary outcomes ongoing adherence to DRS based on attendance for screening following the initial screening post intervention, and economic outcomes.

Authors searched the Cochrane Central Register of Controlled Trials (CENTRAL) and the NHS Economic Evaluation Database (NHS EED) on The Cochrane Library, Ovid MEDLINE, Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE, EMBASE, PsycINFO, the Web of Science Conference Proceedings Citation Index – Science (CPCI-S) (January 1990 to February 2017) and Emerging Sources Citation Index (ESCI), ProQuest Family Health and OpenGrey. Authors also searched the following trials registers: International Standard Randomised Controlled Trial Number (ISRCTN) registry, ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP). Authors did not use any date or language restrictions in the electronic searches. As part of the search strategy also reviewed references of included studies. Two authors independently screened studies for inclusion and extracted data of included studies using a modified version of the EPOC group data collection form.

Studies judged to potentially include economic data were identified and further assessed by an economics reviewer. Data from included economic evaluations were extracted by one reviewer and checked by a second reviewer. Data collection was adapted from the format and guidelines used to produce the structured abstracts of full economic evaluations for inclusion in the NHS EED, which were redesigned to accommodate specific data required for the review. Economic evaluations were classified based on their analytical framework and were coded appropriately.

Two authors independently coded intervention content (QI) as ‘present’ or ‘absent’ for all interventions and control arms. Authors developed an ordered ranking scale to quantify the level of resource needed to deliver each intervention. To determine the feasibility of this approach, authors piloted the scale on a sample of 10 included studies using two members of the study. Each intervention was initially graded between 1 (least resource intensive) and 5 (most resource intensive), or as 0 (unable to determine), with a record of how the reviewer graded each study also provided.

Assessment of included studies was conducted aby two authors independently using the EPOC Group risk-of-bias tools.  Authors report that the identified economic studies were a subset of the studies included in the review, the risk of bias was already assessed. However, assessment of the overall methodological quality of the economic component was still required and was carried out by one reviewer using the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement together with the Consensus on Health Economic Criteria (CHEC). Authors note that the aim of this assessment was to assess the applicability of the scope of each economic evaluation.

In terms of data synthesis, attendance at screening post intervention is a dichotomous outcome and measure of intervention effect was the risk difference (RD), that is, the actual difference in the observed events between experimental and control interventions. For individual RCTs the unit of analysis was the individual participant. For cluster RCTs authors analysed data after adjustment for clustering. In the case of cluster RCTs, when outcomes were presented at the patient level, authors used an established method to adjust for clustering. Authors conducted meta-analyses in Review Manager 5using a random-effects models to estimate RD across studies. Data from patient RCTs and cluster RCTs that were adjusted for clustering were included in the same meta-analyses.

Heterogeneity was assessed by visual inspection of forest plots and by formal statistical tests of heterogeneity (chi-squared test and the I² statistic).

Applicability/external validity:

Overall, authors note that interventions increased DRS and that both DRS-targeted and general QI interventions were effective, particularly when baseline attendance levels were low.

Geographic focus:

Authors note that included studies were all conducted in high-income settings, but 17 studies were conducted in disadvantaged populations. However, authors no dot provide further details.

Summary of quality assessment:

Overall, there is high confidence in the conclusions about the effects of this study. Authors used appropriate methods to ensure that all relevant studies were included in the review avoiding biases. Authors also used appropriate methods to screen studies for inclusion, extract data and assess the methodological quality of included studies. Limitations of included studies were appropriately addressed and no strong policy conclusions were drawn.