Jay Kaplan MD, FACEP, recently argued in ACEP Now (April 2014) that emergency physicians should embrace patient experience metrics because patient satisfaction has been linked to patient adherence to evidence-based recommendations and improved clinical outcomes. However, Dr. Kaplan was selective in his review of the literature and, at times, erroneous. We believe that emergency physicians have legitimate concerns about the potential misuse of patient experience metrics and that Dr. Kaplan’s enthusiasm is unjustified.
Dr. Kaplan concluded that the literature overwhelmingly supports a causal connection between patient satisfaction and clinical care quality, citing a 2013 BMJ Open review.1 However, this review included studies utilizing sophisticated patient communication measures bearing little resemblance to widely used patient experience metrics. Key negative studies were excluded from the review, including a Dartmouth Atlas analysis that found no consistent relationship between satisfaction and clinical care quality.2
Indeed, the Dartmouth Atlas study is consistent with other studies of the relationship between patient satisfaction and technical health care quality.3 While some literature supports an association between patient satisfaction and adherence, patient satisfaction is affected by factors frequently unmeasured in satisfaction studies. In a nationally representative sample, unadjusted positive associations between patient satisfaction and preventive care adherence were eliminated, or even reversed, with sequential adjustment for patient sociodemographics, physical and mental health status, and attitudes toward health care.4
Meanwhile, Dr. Kaplan criticized a study one of us published.5 Within a nationally representative sample, the study found that patients in the highest patient satisfaction quartile (versus the lowest) had 8.8 percent greater total health expenditures, 9.1 percent greater prescription drug expenditures, and significantly higher mortality over a mean follow-up of 3.9 years. The study adjusted for patient-level covariates often not included in prior investigations, including physical and mental health status, chronic illness, and prior health care utilization. The results highlighted the need to better understand the potential link between patient satisfaction and health care utilization, including the use of health care that may, on balance, be harmful.
Dr. Kaplan stated that the study “has no legitimacy” due to three “serious methodological flaws”: 1) that satisfaction was only measured in 2000 and not in later years, 2) that drug and total expenditures were only measured in 2001, and 3) that mortality was assessed in 2001–2006 and never in years when satisfaction or cost were measured. Each statement is false. Regarding the first two, relationships between patient satisfaction and utilization were studied all years from 2000 to 2008. Regarding the third, satisfaction in 2000–2005 and mortality outcomes through 2006 were assessed for the subsample initially enrolled in 2000–2005.