In 2001, an influential report defined quality health care in terms of 6 domains: safety, effectiveness, timeliness, efficiency, patient-centredness and equity. (1) Physicians primarily focus on effectiveness, which can be defined operationally as the degree to which patients receive the recommendations from practice guidelines relevant to their care. Only about 50% of patients receive such recommended care. (2) In addition, patients frequently receive nonrecommended care and undergo tests and treatments with no clear benefits. (3)
In related research, Squires and colleagues provide contemporary estimates of these twin problems of underuse and overuse in Canada. (4) Using Canadian data from 174 studies published from 2007 to 2021, they found a median proportion of inappropriate care (either underuse or overuse) of 30%. This inventory of shortfalls in the delivery of effective care comes after decades of efforts to address such quality problems, (5) begging the question, "What should we be doing to deliver better quality care?"
The mainstream approach to quality improvement--my professional focus for many years--aims to ensure that tests and treatments with established benefits are offered routinely to patients. Yet, efforts to promote the uptake of recommended care have produced few substantive successes. A systematic review of more than 100 interventions to improve professional practice that was published in CMAJ in 1995 concluded that "we have no magic bullets--most interventions achieved small to modest improvements in recommended practices." (6) More than 25 years later, promoting the uptake of effective interventions or de-implementing ineffective ones remains difficult. For instance, 2 common improvement interventions, computerized decision support and performance report cards, typically increase the proportion of patients receiving recommended care by just 5% overall according to the findings of 2 metaanalyses. (7,8) Although some individual trials report larger improvements, the circumstances producing such results remain unclear. Even if worthwhile improvements could be generated consistently, overloading physicians with report cards and reminders for multiple different quality targets is unsustainable.
The authors of the related systematic review (4) identified glycosylated hemoglobin ([HbA.sub.1c]) testing as the most frequently underused diagnostic test in Canada. Addressing underuse of [HbA.sub.1c] testing might prompt use of the many effective therapies for diabetes that have become available in recent years. Yet, more than 100 randomized controlled trials evaluating improvement interventions for outpatients with diabetes achieved an average reduction in [HbA.sub.1c] of just 0.37%. (9) Although future research might identify more effective improvements, the question then becomes how to decide which quality problems to prioritize. This prioritization might depend on the anticipated health benefits from closing any particular shortfall in recommended care. Statins, the most frequent therapeutic example of underuse in the systematic review, (4) provides a...