Article type
Year
Abstract
Background: The US Agency for Healthcare Research and Quality’s (AHRQ) Evidence-based Practice Center (EPC) Program produces systematic reviews to inform healthcare decisions such as recommendations in clinical practice guidelines and national coverage determinations. Reviews also frequently address priorities of healthcare systems to inform initiatives to improve the quality of care delivered. Due to limitations in the literature base, systematic reviews may be inconclusive or only represent a narrow patient population, making it difficult to generalize or apply review findings across healthcare systems.
Objectives: To determine the feasibility, resources, and infrastructure needed to use real-world data from healthcare systems to supplement findings from systematic reviews.
Methods: Three topics recently addressed by AHRQ systematic reviews were chosen for this pilot. A retrospective observational study on each topic was conducted by individual teams using electronic health record (EHR) data from their healthcare system (Mayo Clinic, Children’s Hospital of Philadelphia, and University of Minnesota). One team assessed outcomes of migraine treatment in pregnant persons or those with cardiovascular disease; the second examined clinical characteristics associated with surgical versus nonsurgical treatment of infantile epilepsy; and the third evaluated the use of treatment for osteoporosis after surgery for incident fractures. For each topic, we also examined resource requirements.
Results: We report how patient populations of interest were defined and identified from the EHR and describe the degree to which the outcomes (effect size and strength of evidence) from the analyses of real-world data complement or conflict with the findings from the original systematic review. We also report the feasibility of routinely conducting such supplementary analyses of real-world data, considering additional resource requirements and timeliness.
Conclusions: Our findings suggest that analyses of real-world data from healthcare systems in the United States can supplement systematic review findings by offering additional insights. This is particularly true when decisions involve patient populations such as infants and pregnant women that are seldom included in clinical trials. In our experience, accessing and validating healthcare system data took the most time but may offer efficiencies when standard procedures are established.
Objectives: To determine the feasibility, resources, and infrastructure needed to use real-world data from healthcare systems to supplement findings from systematic reviews.
Methods: Three topics recently addressed by AHRQ systematic reviews were chosen for this pilot. A retrospective observational study on each topic was conducted by individual teams using electronic health record (EHR) data from their healthcare system (Mayo Clinic, Children’s Hospital of Philadelphia, and University of Minnesota). One team assessed outcomes of migraine treatment in pregnant persons or those with cardiovascular disease; the second examined clinical characteristics associated with surgical versus nonsurgical treatment of infantile epilepsy; and the third evaluated the use of treatment for osteoporosis after surgery for incident fractures. For each topic, we also examined resource requirements.
Results: We report how patient populations of interest were defined and identified from the EHR and describe the degree to which the outcomes (effect size and strength of evidence) from the analyses of real-world data complement or conflict with the findings from the original systematic review. We also report the feasibility of routinely conducting such supplementary analyses of real-world data, considering additional resource requirements and timeliness.
Conclusions: Our findings suggest that analyses of real-world data from healthcare systems in the United States can supplement systematic review findings by offering additional insights. This is particularly true when decisions involve patient populations such as infants and pregnant women that are seldom included in clinical trials. In our experience, accessing and validating healthcare system data took the most time but may offer efficiencies when standard procedures are established.