Retail Clinic Visits For Low-Acuity Conditions Increase Utilization And Spending

Retail clinics are clinics located in pharmacies, grocery stores, and “big-box stores” such as Walmart and Target. The clinics primarily provide care for a limited set of low-acuity acute conditions such as urinary tract infections and sinusitis as well as preventive services such as immunizations. ^1–4 Evening and weekend hours, short wait times, and the ability to provide a walk-in visit may make retail clinics more convenient for patients than visits to physician offices or emergency departments (EDs). ⁵

There are now almost two thousand retail clinics in the United States, and they receive more than six million patient visits per year. ^6,7 Compared to similar visits to physician offices and EDs, retail clinic visits are less costly for both patients and health plans. ^3,8 As a result, retail clinics have been viewed by policy makers and insurers as a method to decrease health care spending. ^9,10

Whether retail clinics actually decrease spending is unknown, since—to our knowledge—the impact of retail clinics on utilization has never been assessed. Retail clinic visits may save money by substitution (replacing more expensive visits to physician offices or EDs). However, they may also lead to new utilization (when patients who previously would not have sought care visit a retail clinic because of its lower price or perceived convenience), which increases both the total number of patient visits and spending.

In this study we assessed whether retail clinic visits for low-acuity conditions represented substitution or new utilization, using data from one large health insurer across twenty-two US cities.

Study Data And Methods

Overview

We used a matched design to track the utilization and spending associated with eleven low-acuity conditions among two sets of Aetna enrollees, users and nonusers of retail clinics, in twenty-two cities for the period 2010–12. We defined the two sets of enrollees by where they sought care after 2010. Retail clinic users were those who used a retail clinic for the first time in 2011–12. Nonusers constituted a matched population who used another type of care site in 2011–12 for a similar low-acuity condition.

Because of how we identified the set of retail clinic users, by definition the use of retail clinics increased in that group during the study period. We determined whether retail clinics represented substitution or new utilization by focusing on the change in use of other care sites (physician offices and EDs) relative to the change in use of retail clinics.

We focused on visits for eleven low-acuity conditions commonly seen at retail clinics. ⁶ These conditions accounted for 62.3 percent of all visits to retail clinics in 2012. For International Classification of Diseases , Ninth Revision (ICD-9), diagnosis codes for these conditions, see the online Appendix. ¹¹

Data Source

We used claims and enrollment data for the period 2010–12 for 13.3 million Aetna enrollees in twenty-two cities with retail clinics. Aetna covered retail clinic visits during the entire study period, and copayments for those visits were comparable to copayments for physician office visits.

We excluded enrollees over age sixty-five because they were likely to have Medicare, and some of their claims might not have been captured in our data. We also excluded enrollees living in a ZIP code that was more than twenty miles from any retail clinic in 2012, because previous research suggests that people who live at this distance from a clinic make little use of it. ⁶

Tracking Use Of Retail Clinics

We tracked quarterly use of each retail clinic for low-acuity conditions per 100 enrollees in the twenty-two cities. We reviewed the provider files by hand to ensure that we identified all retail clinics.

Quarterly use of retail clinics for low-acuity conditions was highly variable and increased during the winter months (see Appendix Exhibit A1). ¹¹ To facilitate the visualization of underlying time trends, we adjusted for seasonal fluctuation by presenting a rolling average over four quarters.

Defining Study Cohorts

Aetna provided data for all enrollees with at least one retail clinic visit at any time in the period 2010–12 ( $n=519,542$ ) and a random sample of other enrollees ( $n=861,557$ ).

We identified all enrollees who had a visit for a low-acuity condition at any point in the period 2011–12. Enrollees who had a visit to a retail clinic before 2011 were excluded from the analysis. We divided the remaining enrollees into two cohorts based on where they sought care for the low-acuity condition: those who visited a retail clinic and those who did not.

From the nonuser cohort, we selected a subset that matched the retail clinic user cohort on patient characteristics such as age, sex, location, and health status (for more details, see the Appendix). ¹¹ We used propensity score matching ^12,13 to find two other Aetna enrollees similar to each retail clinic user. This allowed us to minimize selection bias by comparing populations with similar propensities to use retail clinics.

Measuring Use And Spending

We determined whether retail clinic visits represented substitution or new utilization by focusing on the change in use of sites other than retail clinics relative to the use of retail clinics. Therefore, our measure of utilization was the number of average annual visits to a physician office and ED per 100 people. The numerator was the number of visits to physician offices and EDs for a low-acuity condition in the previous year. The denominator was all members of a cohort continuously enrolled for the year.

Our spending measure was the average annual per person spending on visits for a low-acuity condition. This measure combined the increase in spending from new utilization and the savings from substitution. Previous work found visits to retail clinics to be less expensive than physician office or ED visits because the amount spent on the visit itself is lower and because retail clinics perform fewer tests than physician offices or EDs (follow-up visit rates are similar). ^3,8 Therefore, to accurately estimate the savings due to substitution, we measured the spending for each visit and all other services, including tests, on the visit day and the two following days. In sensitivity analyses, we focused just on spending for the visit itself and varied the time window from one to five days after the visit (for more details, see the Appendix). ¹¹

Statistical Analyses

To estimate the changes in physician office and ED utilization and spending associated with the use of retail clinics, we compared the levels of utilization and spending in the period 2010–12 for both sets of enrollees. We used multiple linear regression models with one observation for each combination of year, city, and set of enrollees (retail clinic users or nonusers). To address any selection bias that might remain after propensity score matching, we used difference-in-differences models. ^14–16

Our independent variables were indicators for the cohort of enrollees (retail clinic users or nonusers), year, the interaction of the cohort and year, and each city. We compared changes in our outcomes from 2010 to 2012 for the two cohorts. The differential change for the user cohort relative to the nonuser cohort, captured by the interaction of user cohort and year, serves as an estimate of the change in visits to physician offices and EDs.

To estimate what fraction of retail clinic visits represented substitution as opposed to new utilization, we compared the decrease in physician office and ED visits to the increase in retail clinic visits. We used SAS, version 9.22, for all analyses.

Sensitivity Analyses

In a sensitivity analysis to assess whether our results would be similar for a different time period, we conducted the same analysis but compared users and nonusers of retail clinics in the period 2005–09. For details regarding the sensitivity analyses, see the Appendix. ¹¹ In a second analysis, we compared utilization for the eleven low-acuity conditions among Aetna enrollees who lived close to a retail clinic versus utilization for those who did not.

To address concerns that retail clinic users may have increased utilization for all conditions, we conducted falsification tests. We examined trends in musculoskeletal strains—common low-acuity conditions that are addressed by physician offices and EDs but not retail clinics—among people in the user cohort versus those in the nonuser cohort. We also explored differences by age, by conducting our analyses for children and adults separately.

Limitations

Our study had several limitations. First, our measure of physician office visits included urgent care center visits: We could not distinguish the two sites of care in the data. We also could not identify e-visits, direct-to-consumer telemedicine visits, or use of nurse health lines, all of which are methods by which patients seek care for low-acuity conditions.

Second, we focused only on changes in utilization and spending for a set of low-acuity conditions commonly managed by retail clinics. We could not assess the impact of retail clinics on overall utilization and total health care spending because we did not have data on inpatient care or pharmaceutical utilization—two large components of health care spending. We focused only on medical costs and could not estimate changes in nonmedical costs, such as time missed from work. Incorporating these indirect costs, particularly for EDs, which often have long wait times, would increase the relative benefit of retail clinics from employers’ or patients’ perspectives.

Third, the fraction of retail clinic visits that represented new utilization could change in the future. Our study included only people with commercial insurance, and changes associated with the use of retail clinics among the uninsured or people with government insurance might differ.

Fourth, our results could have been affected by selection bias. However, that concern was reduced because we found similar utilization patterns for users and nonusers of retail clinics before the clinics opened, and because of the results of our falsification tests.

Fifth, unlike retail clinics, physicians might have addressed issues (for example, hypertension) other than the low-acuity problems that led to a visit. If that was the case, we may have overestimated the savings from the substitution of retail clinic visits for physician visits and underestimated how much the use of retail clinics increased spending.

Lastly, our data did not allow us to assess the extent to which retail clinics may improve access to care. We consider this a limitation because many retail clinic users do not have a usual source of care. However, we did perform an additional sensitivity analysis in which we limited the sample to people who did not visit a retail clinic in 2010—a population that likely included people who did not have a regular source of care. The results of that analysis are presented in Appendix Exhibit A2 ¹¹ and briefly discussed below in the “Discussion” section.

Study Results

Retail clinics first opened in the twenty-two markets we studied between early 2006 and late 2007. Across all enrollees, use of retail clinics increased from no visits per thousand people per quarter in 2005 to eleven visits per thousand people in the fourth quarter of 2012 ( Exhibit 1 ). In 2012, 3.0 percent of all enrollees used a retail clinic (data not shown).

SOURCE Authors’ analysis of 2005–12 Aetna claims and enrollment data. NOTES The scatterplot data represent visits to a retail clinic in a given quarter for a low-acuity condition (defined in the text). There was greater use of retail clinics in winter months and lower use in summer months. To account for this seasonal variation, the line represents the rolling average of retail clinic visits over the previous four quarters.

Of the 13.3 million Aetna enrollees, 3.0 million had at least one visit for a low-acuity condition in the period 2011–12. These enrollees were divided into retail clinic users (0.4 million) and nonusers (2.6 million). Before propensity score matching, compared to nonusers, users were more likely to be younger, female, and healthier; live closer to retail clinics; and live in higher-income ZIP codes. After matching, the populations were similar on all observed characteristics (Appendix Exhibit A3). ¹¹

Utilization, 2010–12

After matching, both retail clinic users and nonusers had seventy-one combined physician office and ED visits for low-acuity conditions per hundred people in 2010 ( Exhibit 2 ). From 2010 to 2012 the user cohort had an increase of 69 retail clinic visits for low-acuity problems per 100 enrollees (95% confidence interval: 66, 72). Compared to nonusers, the users of retail clinics had a reduction in the number of physician office and ED visits combined (29 visits per 100 enrollees; 95% CI: −46, −12).

To determine whether retail clinic visits represented substitution or new utilization, we compared the change in visits to physician offices and EDs to the change in visits to retail clinics. We estimated that 42 percent of the retail clinic visits represent substitution and 58 percent represent new utilization ( Exhibit 3 ). Of the visits that represent substitution, we estimated that 93.1 percent represent substitution of physician office visits and 6.9 percent represent substitution of ED visits (data not shown).

SOURCE Authors’ analysis of 2010–12 Aetna claims and enrollment data. NOTES Low-acuity conditions are defined in the text. Expected visits were based on utilization trends among nonusers of retail clinics. ED is emergency department.

Spending, 2010–12

Per person per year spending on visits for low-acuity conditions among retail clinic users increased by $35 (95% CI: 34, 37) for retail clinic visits and decreased by $21 (95% CI: −31, −11) for physician office and ED visits, compared to spending among nonusers ( Exhibit 4 ), from 2010 to 2012. Overall low-acuity condition spending increased by $14 (95% CI: 5, 23). Thus, the savings from the substitution of retail clinic visits for physician office and ED visits ($21) offset 60 percent of the increased spending on retail clinic visits ($35).

In falsification tests, we observed no difference between the two cohorts’ utilization trends for visits for musculoskeletal strains (Appendix Exhibit A4). ¹¹ When we used a different time period (2005–09), we had substantively similar results for the fraction of retail clinic visits that represent new utilization versus substitution (Appendix Exhibit A5), ¹¹ We also had similar results in a comparison of children and adults (Appendix Exhibit A6). ¹¹ This was also true when we used a different analytic method—comparing visits to retail clinics among those who lived near or far from a retail clinic—though the differences in this analysis were not statistically significant (Appendix Exhibit A5). ¹¹

Discussion

Retail clinics have been viewed by policy makers and insurers as a way to decrease health care spending. The impact of retail clinics on this spending depends on whether patients visit a retail clinic instead of a physician office or ED (substitution) or visit a retail clinic when they otherwise would not seek care (new utilization). We found that roughly two-fifths of retail clinic visits for low-acuity conditions represented substitution, while the other three-fifths represented new utilization. In total, the increased spending from new utilization outweighed the savings from substitution. Instead of decreasing spending overall, we found that use of a retail clinic was associated with 21 percent higher spending for low-acuity conditions (an increase of $14 per person per year relative to $66 spent by nonusers).

It may not be surprising that retail clinics increase utilization for low-acuity conditions. Because retail clinics focus on convenience and lower per visit costs, some experts have called them a disruptive innovation in health care. ¹⁷ The increased utilization and spending we observed is consistent with innovations in other industries and other areas of health care. For example, the innovation of personal computers increased the number of computers sold, ¹⁸ and the introduction of laparoscopic technology greatly increased the number of cholecystectomies performed. ¹⁹ In each of these cases, the relative advantage of the innovation increased the number of potential users, thereby increasing utilization and, subsequently, spending.

Whether the increase in utilization and spending we observed is judged to be good or bad depends on one’s perspective. Some experts have defined value as health outcomes per dollar spent and have emphasized that value should be judged from the patient’s perspective. ²⁰ It is reasonable to assume that patients who seek care at retail clinics gain value from new utilization if they would otherwise have not sought care for their health conditions. The observed increase in utilization might also represent an unmet demand by patients who previously lacked a source of care. Our sensitivity analysis of infrequent health care users suggested that substitution may have been more common among those without a usual source of care than those with a regular primary care physician (Appendix Exhibit A2). ¹¹ In theory, increased use of retail clinics could prevent costly complications by addressing an illness earlier in its course than would otherwise have been the case. Given these potential benefits, patients might be willing to pay out of pocket for retail clinic care.

However, the majority of care at retail clinics is paid for by health plans, including Medicare. ^4,21 These payers may not judge the increase in utilization as increasing value because the low-acuity health issues addressed at retail clinics are typically self-limited, and therefore a visit may not improve health in the long term. ^22–24

Judgment on the value of retail clinic visits is also influenced by the clinics’ primary focus on low-acuity conditions. Immunizations and preventive care (for example, sports and school physicals, blood pressure checks, and cholesterol screenings) represent up to 40 percent of retail clinic visits, and retail clinics have only recently begun offering chronic illness care. ^2,6,25 Retail clinic patients tend to be young and include those who lack a usual source of care, ^2,6,25 a population at higher risk than those with a usual source of care for low rates of preventive care ^26,27 and poor chronic disease management. ²⁸ If retail clinics also increase the use of preventive care or chronic disease care, this would likely be viewed as improving value because these are aspects of health care perceived to be underused. ^29,30

It is important to emphasize that the relative increase in patient and health plan spending on low-acuity conditions driven by retail clinics is quite small ($14 per person per year). Furthermore, only 3 percent of all Aetna enrollees in the markets we studied visited a retail clinic in 2012. In addition, annual per person spending on the low-acuity conditions we studied ($35 in retail clinics and $46–$66 in physician offices and EDs) represented a tiny sliver of overall health care spending, given that per capita spending on health care in the United States in 2012 was $8,927. ³¹

Policy Implications

A key finding from our study—that spending increased for retail clinic visits for low-acuity conditions—seems to refute the notion that retail clinics save patients and health plans money by replacing higher-cost physician office and ED visits for those conditions. ^9,10 Our findings can therefore help inform health plans’ and public payers’ coverage decisions for retail clinics.

For example, almost all health plans include retail clinics in their provider networks. Several health plans have created financial incentives to encourage the use of such clinics by waiving copays for clinic visits, and others have used direct marketing to encourage enrollees to use them. If these interventions were driven by a desire to decrease health care spending, our results imply that they may not be completely effective.

We believe that our findings can also inform the many other efforts to improve access to and convenience in health care. Physician offices and EDs are adopting strategies to increase access and decrease waiting times. ^32,33 For example, some physicians offer extended hours and open-access scheduling, ³² and others are adopting characteristics of patient-centered medical homes, for which one key goal is to improve access. ³⁴ Health care systems are beginning to focus on increasing convenience by providing care through their own retail clinics, ^17,32 e-visits, ^35–37 and telemedicine. ^38,39 The improved convenience of these initiatives may increase utilization and spending.

We found that approximately two-fifths of retail clinic visits substituted for physician office or ED visits, a finding with both positive and negative consequences. As noted above, this substitution lowered spending for health plans and patients and offset more than half of the increase in spending that arose from increased utilization. Fewer office visits could improve efficiency in the health care delivery system by freeing physicians to handle more complicated cases that are more appropriate for their higher level of training than simpler cases, and thereby potentially lessen the primary care shortage in the United States. ⁴⁰ However, retail clinics have been found to decrease continuity of care and therefore may disrupt the physician-patient relationship. ^41,42 This may be of particular concern for people with multiple chronic illnesses, for whom continuity of care is critical.

Conclusion

Though retail clinics have been promoted as a means of reducing health care spending by substituting for more expensive providers, we found that most retail clinic visits represented new utilization and therefore increased health care spending per person per year for low-acuity conditions. These results should help inform payers’ coverage decisions for retail clinics and other care options that increase convenience and access. Future research should investigate how retail clinics affect the coordination of care, care for chronic illnesses, and overall spending.

ACKNOWLEDGMENTS

These results were previously presented at the AcademyHealth Annual Research Meeting, Baltimore, Maryland, June 2013. The authors gratefully acknowledge the Robert Wood Johnson Foundation for financial support of this work and Aetna for providing claims data and helpful comments on the manuscript. The Robert Wood Johnson Foundation had no role in the design or conduct of the study; no role in the collection, management, analysis, or interpretation of the data; no role in the preparation, review, or approval of the manuscript; and no role in the decision to submit the manuscript for publication.

NOTES

• 1 Scott MK . Health care in the express lane: retail clinics go mainstream [Internet]. Oakland (CA) : California HealthCare Foundation ; 2007 Sep [cited 2016 Jan 26 ]. Available from: http://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20H/PDF%20HealthCareInTheExpressLaneRetailClinics2007.pdf Google Scholar
• 2 Mehrotra A , Wang MC , Lave JR , Adams JL , McGlynn EA . Retail clinics, primary care physicians, and emergency departments: a comparison of patients’ visits . Health Aff (Millwood) . 2008 ; 27 ( 5 ): 1272 – 82 . Go to the article, Google Scholar
• 3 Mehrotra A , Liu H , Adams JL , Wang MC , Lave JR , Thygeson NM , et al. Comparing costs and quality of care at retail clinics with that of other medical settings for 3 common illnesses . Ann Intern Med . 2009 ; 151 ( 5 ): 321 – 8 . Crossref, Medline, Google Scholar
• 4 Rudavsky R , Pollack CE , Mehrotra A . The geographic distribution, ownership, prices, and scope of practice at retail clinics . Ann Intern Med . 2009 ; 151 ( 5 ): 315 – 20 . Crossref, Medline, Google Scholar
• 5 Wang MC , Ryan G , McGlynn EA , Mehrotra A . Why do patients seek care at retail clinics, and what alternatives did they consider? Am J Med Qual . 2010 ; 25 ( 2 ): 128 – 34 . Crossref, Medline, Google Scholar
• 6 Ashwood JS , Reid RO , Setodji CM , Weber E , Gaynor M , Mehrotra A . Trends in retail clinic use among the commercially insured . Am J Manag Care . 2011 ; 17 ( 11 ): e443 – 8 . Medline, Google Scholar
• 7 Mehrotra A , Lave JR . Visits to retail clinics grew fourfold from 2007 to 2009, although their share of overall outpatient visits remains low . Health Aff (Millwood) . 2012 ; 31 ( 9 ): 2123 – 9 . Go to the article, Google Scholar
• 8 Thygeson M , Van Vorst KA , Maciosek MV , Solberg L . Use and costs of care in retail clinics versus traditional care sites . Health Aff (Millwood) . 2008 ; 27 ( 5 ): 1283 – 92 . Go to the article, Google Scholar
• 9 Eibner CE , Hussey PS , Ridgely MS , McGlynn EA . Controlling health care spending in Massachusetts: an analysis of options [Internet]. Santa Monica (CA) : RAND Health ; 2009 Aug [cited 2016 Jan 26 ]. Available from: http://www.rand.org/content/dam/rand/pubs/technical_reports/2009/RAND_TR733.pdf Google Scholar
• 10 Hussey PS , Eibner C , Ridgely MS , McGlynn EA . Controlling U.S. health care spending—separating promising from unpromising approaches . N Engl J Med . 2009 ; 361 ( 22 ): 2109 – 11 . Crossref, Medline, Google Scholar
• 11 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 12 Rosenbaum PR , Rubin DB . Constructing a control group using a multivariate matched sample methods that incorporate propsensity score . Am Stat . 1985 ; 39 ( 1 ): 33 – 8 . Google Scholar
• 13 Rosenbaum PR . Observational studies . 2nd ed . New York (NY) : Springer-Verlag ; 2002 . Crossref, Google Scholar
• 14 Bang H , Robins JM . Doubly robust estimation in missing data and causal inference models . Biometrics . 2005 ; 61 ( 4 ): 962 – 73 . Crossref, Medline, Google Scholar
• 15 Kang JDY , Schafer JL . Demystifing double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data . Stat Sci . 2007 ; 22 ( 4 ): 523 – 39 . Crossref, Google Scholar
• 16 Dimick JB , Ryan AM . Methods for evaluating changes in health care policy: the difference-in-differences approach . JAMA . 2014 ; 312 ( 22 ): 2401 – 2 . Crossref, Medline, Google Scholar
• 17 Bohmer R . The rise of in-store clinics—threat or opportunity? N Engl J Med . 2007 ; 356 ( 8 ): 765 – 8 . Crossref, Medline, Google Scholar
• 18 Christensen CM , Bohmer RMJ , Kenagy J . Will disruptive innovations cure healthcare? Harv Bus Rev . 2000 ; 78 ( 5 ): 102 – 17 . Medline, Google Scholar
• 19 Nenner RP , Imperato PJ , Rosenberg C , Ronberg E . Increased cholecystectomy rates among Medicare patients after the introduction of laparoscopic cholecystectomy . J Community Health . 1994 ; 19 ( 6 ): 409 – 15 . Crossref, Medline, Google Scholar
• 20 Porter ME . What is value in health care? N Engl J Med . 2010 ; 363 ( 26 ): 2477 – 81 . Crossref, Medline, Google Scholar
• 21 Martin AB , Lassman D , Washington B , Catlin A National Health Expenditure Accounts Team . Growth in US health spending remained slow in 2010; health share of gross domestic product was unchanged from 2009 . Health Aff (Millwood) . 2012 ; 31 ( 1 ): 208 – 19 . Go to the article, Google Scholar
• 22 Garbutt JM , Banister C , Spitznagel E , Piccirillo JF . Amoxicillin for acute rhinosinusitis: a randomized controlled trial . JAMA . 2012 ; 307 ( 7 ): 685 – 92 . Crossref, Medline, Google Scholar
• 23 Smucny J , Fahey T , Becker L , Glazier R . Antibiotics for acute bronchitis . Cochrane Database Syst Rev . 2004 ( 4 ): CD000245 . Google Scholar
• 24 Del Mar C , Glasziou P , Hayem M . Are antibiotics indicated as initial treatment for children with acute otitis media? A meta-analysis . BMF . 1997 ; 314 ( 7093 ): 1526 – 9 . Crossref, Medline, Google Scholar
• 25 Frieden J . Retail clinics branch into chronic disease treatment . ABC News [serial on the Internet]. 2010 Mar 25 [cited 2016 Jan 26 ]. Available from: http://abcnews.go.com/Health/HealthCare/retail-clinics-branch-chronic-disease-treatment/story?id=10152224 Google Scholar
• 26 Bindman AB , Grumbach K , Osmond D , Vranizan K , Stewart AL . Primary care and receipt of preventive services . J Gen Intern Med . 1996 ; 11 ( 5 ): 269 – 76 . Crossref, Medline, Google Scholar
• 27 Ettner SL . The timing of preventive services for women and children: the effect of having a usual source of care . Am J Public Health . 1996 ; 86 ( 12 ): 1748 – 54 . Crossref, Medline, Google Scholar
• 28 Gregg EW , Karter AJ , Gerzoff RB , Safford M , Brown AF , Tseng CW , et al. Characteristics of insured patients with persistent gaps in diabetes care services: the Translating Research into Action for Diabetes (TRIAD) study . Med Care . 2010 ; 48 ( 1 ): 31 – 7 . Crossref, Medline, Google Scholar
• 29 DeVoe JE , Fryer GE , Phillips R , Green L . Receipt of preventive care among adults: insurance status and usual source of care . Am J Public Health . 2003 ; 93 ( 5 ): 786 – 91 . Crossref, Medline, Google Scholar
• 30 McGlynn EA , Asch SM , Adams J , Keesey J , Hicks J , DeCristofaro A , et al. The quality of health care delivered to adults in the United States . N Engl J Med . 2003 ; 348 ( 26 ): 2635 – 45 . Crossref, Medline, Google Scholar
• 31 Martin AB , Hartman M , Benson J , Catlin A National Health Expenditure Accounts Team . National health spending in 2014: faster growth driven by coverage expansion and prescription drug spending . Health Aff (Millwood) . 2016 ; 35 ( 1 ): 150 – 60 . Go to the article, Google Scholar
• 32 Murray M , Berwick DM . Advanced access: reducing waiting and delays in primary care . JAMA . 2003 ; 289 ( 8 ): 1035 – 40 . Crossref, Medline, Google Scholar
• 33 Wiler JL , Gentle C , Halfpenny JM , Heins A , Mehrotra A , Mikhail MG , et al. Optimizing emergency department front-end operations . Ann Emerg Med . 2010 ; 55 ( 2 ): 142 – 60 . Crossref, Medline, Google Scholar
• 34 Starfield B . Primary care: balancing health needs, services, and technology . Rev. ed . New York (NY) : Oxford University Press ; 1998 . Google Scholar
• 35 Tang PC , Lee TH . Your doctor’s office or the Internet? Two paths to personal health records . N Engl J Med . 2009 ; 360 ( 13 ): 1276 – 8 . Crossref, Medline, Google Scholar
• 36 Rohrer JE , Angstman KB , Adamson SC , Bernard ME , Bachman JW , Morgan ME . Impact of online primary care visits on standard costs: a pilot study . Popul Health Manag . 2010 ; 13 ( 2 ): 59 – 63 . Crossref, Medline, Google Scholar
• 37 Courneya PT , Palattao KJ , Gallagher JM . HealthPartners’ online clinic for simple conditions delivers savings of $88 per episode and high patient approval . Health Aff (Millwood) . 2013 ; 32 ( 2 ): 385 – 92 . Go to the article, Google Scholar
• 38 Perednia DA , Allen A . Telemedicine technology and clinical applications . JAMA . 1995 ; 273 ( 6 ): 483 – 8 . Crossref, Medline, Google Scholar
• 39 Mehrotra A , Paone S , Martich GD , Albert SM , Shevchik GJ . A comparison of care at e-visits and physician office visits for sinusitis and urinary tract infection . JAMA Intern Med . 2013 ; 173 ( 1 ): 72 – 4 . Crossref, Medline, Google Scholar
• 40 Bodenheimer T , Pham HH . Primary care: current problems and proposed solutions . Health Aff (Millwood) . 2010 ; 29 ( 5 ): 799 – 805 . Go to the article, Google Scholar
• 41 Berman S . Continuity, the medical home, and retail-based clinics . Pediatrics . 2007 ; 120 ( 5 ): 1123 – 5 . Crossref, Medline, Google Scholar
• 42 Reid RO , Ashwood JS , Friedberg MW , Weber ES , Setodji CM , Mehrotra A . Retail clinic visits and receipt of primary care . J Gen Intern Med . 2013 ; 28 ( 4 ): 504 – 12 . Crossref, Medline, Google Scholar