Article type
Year
Abstract
Background: Patients with multiple chronic conditions are among the largest consumers of healthcare and face substantial health disparities. Randomized controlled trials have not sufficiently studied these populations, while observational studies face the problem of confounding by indication, creating a dual problem for external and internal validity, respectively. In particular, statins are among the most prescribed treatments to prevent heart attacks and strokes among patients with combinations of diabetes mellitus type II, coronary artery disease and chronic kidney disease, yet the impact of having different combinations of these conditions on the effectiveness of statin therapy remains unclear.
Objectives: To use the case of preventive statin therapy to determine whether systematic review and meta-analysis can facilitate the calculation of the impact of having multiple chronic conditions on the effectiveness of healthcare interventions.
Methods: We performed a systematic review to find, assess and synthesize all relevant research studies. We used meta-analysis to generate pooled effect sizes for studies with populations that had different combinations of chronic conditions. We used meta-regression and subgroup analysis to control for study-level characteristics, including risks of bias and demographics.
Results: We identified 9962 articles from the search, and included 90 research studies. Through this study we found that having multiple chronic conditions significantly reduces the effectiveness of primary and secondary preventive statin therapy for both myocardial infarction and stroke, and use of a de novo study-level index of confounding also showed that the impact of confounding by indication can be controlled via meta-analysis.
Conclusions: Systematic review and meta-analysis can help control for confounding by indication, and facilitate use of the full body of comparative effectiveness evidence to study the impact of multiple chronic conditions. We present our methodological and statistical approaches to accomplish this objective, using systematic review and meta-analysis, as tools to help move evidence quickly into clinical practice for these populations with special care needs.
Patient or healthcare consumer involvement: Patients were not involved in the work to date, as the approach at this stage has been purely methodological.
Objectives: To use the case of preventive statin therapy to determine whether systematic review and meta-analysis can facilitate the calculation of the impact of having multiple chronic conditions on the effectiveness of healthcare interventions.
Methods: We performed a systematic review to find, assess and synthesize all relevant research studies. We used meta-analysis to generate pooled effect sizes for studies with populations that had different combinations of chronic conditions. We used meta-regression and subgroup analysis to control for study-level characteristics, including risks of bias and demographics.
Results: We identified 9962 articles from the search, and included 90 research studies. Through this study we found that having multiple chronic conditions significantly reduces the effectiveness of primary and secondary preventive statin therapy for both myocardial infarction and stroke, and use of a de novo study-level index of confounding also showed that the impact of confounding by indication can be controlled via meta-analysis.
Conclusions: Systematic review and meta-analysis can help control for confounding by indication, and facilitate use of the full body of comparative effectiveness evidence to study the impact of multiple chronic conditions. We present our methodological and statistical approaches to accomplish this objective, using systematic review and meta-analysis, as tools to help move evidence quickly into clinical practice for these populations with special care needs.
Patient or healthcare consumer involvement: Patients were not involved in the work to date, as the approach at this stage has been purely methodological.