Background: Few studies have examined the association between patient-initiated electronic messaging (e-messaging) and clinical outcomes in fee-for-service settings.
Objective: To estimate the association between patient-initiated e-messages and quality of care among patients with diabetes and hypertension.
Design: Longitudinal observational study from 2009 to 2013. In March 2011, the medical group eliminated a $60/year patient user fee for e-messaging and established a provider payment of $3–5 per patient-initiated e-message. Quality of care for patients initiating e-messages was compared before and after March 2011, relative to nonmessaging patients. Propensity score weighting accounted for differences between e-messaging and nonmessaging patients in generalized estimating equations.
Setting: Large multispecialty practice in California compensating providers’ fee-for-service.
Subjects: Patients with diabetes (N=4232) or hypertension (N=15,463) who had activated their online portal but not e-messaged before e-messaging became free.
Measures: Quality of care included HEDIS-based process measures for hemoglobin (Hb) A1c, blood pressure, low-density lipoprotein (LDL), nephropathy, and retinopathy tests, and outcome measures for HbA1c, blood pressure, and LDL. E-messaging was measured as counts of patient-initiated e-message threads sent to providers. Patients were categorized into quartiles by e-messaging frequency.
Results: The probability of annually completing indicated tests increased by 1%–7% for e-messaging patients, depending on the outcome and e-messaging frequency. E-messaging was associated with small improvements in HbA1c and LDL for some patients with diabetes.
Conclusion: Patient-initiated e-messaging may increase the likelihood of completing recommended tests, but may not be sufficient to improve clinical outcomes for most patients with diabetes or hypertension without additional interventions.
*Abt Associates, Cambridge, MA
†Qualis Health, Seattle, WA
‡Sutter Health, Emeryville, CA
§Palo Alto Medical Foundation Research Institute, Palo Alto, CA
Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website, www.lww-medicalcare.com.
Presented at 3 meetings in the past: The AcademyHealth Annual Research Meeting Health Information Technology Interest Group, San Diego, CA, June 7, 2014; the American Medical Informatics Association Annual Symposium, Washington, DC, November 17, 2014; and the HMO Research Network Conference, Long Beach, CA, March 13, 2015.
Supported by Agency of Healthcare Research and Quality (R18 HS019167; PI: M.T.-S.) and the Levy Family Foundation for financial support.
S.R.M. was a Post-Doctoral Fellow and Laura Panattoni was a Research Economist at Palo Alto Medical Foundation Research Institute during the time the work for this study was completed. The remaining authors declare no conflict of interest.
Reprints: Sean R. McClellan, PhD, Abt Associates, 55 Wheeler Street, Cambridge, MA 02138. E-mail: sean_mcclellan@abtassoc.com.
