Electronic Visits for Minor Acute Illnesses: Analysis of Patient Demographics,Prescription Rates,and Follow-Up Care Within an Asynchronous Text-Based Online Visit

Introduction

Healthcare delivery is a constantly changing domain. As patients ask for increased access to healthcare advice, more cost-effective and less expensive options for care, and more convenient models of treatment, many novel delivery models have evolved. Advances in technology have increased patient's electronic interactions with healthcare. Secure patient portals allow patients 24/7 access to their medical records, allowing patients to view their own test results, request prescription refills, and interact asynchronously with their healthcare providers. ¹

Electronic visits (eVisits) using secure messaging and patient online portals are one new way of providing healthcare for minor acute illnesses, allowing patients to communicate electronically with a healthcare provider asynchronously for diagnosis and treatment of symptoms. eVisits are increasingly being utilized by healthcare institutions with multiple goals. eVisits have the potential to improve access to healthcare by providing options for care on weekends, holidays, and nights when emergency department care may be the only other alternative. Additionally, they offer patients the ability to be treated from home or work, avoiding long waits in an acute care setting and potentially decreasing disease transmission among patients in waiting rooms. This is achieved at a reduced cost when compared with traditional care. ² Despite the increasing popularity of eVisits, research to establish this type of care delivery as a cost-effective and safe alternative to traditional face-to-face (F2F) care is sparse. ^{3 –7} In addition to cost savings, a recent study by Tan et al. suggests that virtual visits may actually have better outcomes (lower rates of follow-up) for minor acute illness, such as upper respiratory illness, when compared directly with F2F visits. ⁸ A potential benefit of provider–patient messaging and eVisits could be a decrease in F2F provider visits; however, studies have shown mixed results of the impact of patient–provider electronic communication on primary care F2F visits and telephone call contacts. ^{7,9 –12} Concerns regarding patient–provider messaging have included the possibility of increased burden for the provider by creating work not captured within the billing and documentation of a traditional F2F visit ¹³ as well as care provided outside of an already established patient–clinician relationship.

We conducted a retrospective record review of patients who received care through an eVisit for a minor acute illness and reviewed 30-day outcomes of follow-up, emergency department visits, hospitalizations, and death for the same or related conditions.

Methods

We retrospectively analyzed charts from all eVisit requests by Mayo Clinic Rochester primary care empaneled patients that occurred between March 1, 2015, and December 6, 2015. This time frame was chosen because it coincided with the launch of a restructured online visit process aimed at improving efficiency, consistency, and response time for eVisits. Documentation in the Mayo Clinic EMR was manually reviewed to determine whether the follow-up was related to the eVisit concern. Thirty-day outcomes evaluated included primary care contact by phone call, portal message, or F2F visit; F2F at a Mayo Clinic retail-based acute care clinic (Express Care); emergency department visits; hospitalization; and/or death for the same or related symptom/condition indicated in the original eVisit. Additional data collected included patient demographics, time of eVisit (during clinic hours or not), whether a prescription medication was provided at the time of the eVisit, and if the patient was specifically instructed by the eVisit provider to follow-up for a F2F visit. Follow-up visits at an outside institution would not have been captured, but based on limited alternative options for healthcare in the area, we anticipate this number would be low.

Comparison of outcomes was performed using a chi-square test for categorical data and a t test for continuous variables. JMP, version 10.0 (SAS Institute, Inc., Cary, NC), was used for statistical analysis.

Patients were excluded from the study if they requested an eVisit for a symptom or diagnosis not part of the restricted diagnosis menu. eVisits for sore throat were also excluded as the algorithmic questions used frequently resulted in incomplete data collection and thus more frequent referrals for in-person examinations due to inadequate data. This technical issue persisted for the duration of the study. Additionally, part of the treatment protocol for sore throats includes referring the patients to a designated site for collection of a throat swab, thus possibly increasing the appearance of follow-up frequency. This study was approved by the Mayo Clinic institutional review board.

Results

During the time period studied, there were 1,199 eVisit requests. Of these, 138 (12%) were for pharyngitis and 52 (4%) were for conditions not covered by eVisit categories. Some examples of requests for eVisits not part of the restrictive menu included symptoms of abdominal pain, diarrhea, thrush, tongue pain, hair loss, diabetes, skin, ear pain, and patients who were outside the appropriate age range for a listed condition. After excluding these, there were 1,009 eVisits to analyze. Of these, 708 (70%) were requested during clinic hours. The average age of eVisit patients was 36.6 years (range 1.5–73 years) with patients <18 years of age representing 7% of eVisits and patients 60 years and older representing 3% of eVisits. Female patients accounted for 88% of eVisits. The frequency of each diagnosis request in an eVisit is shown in Table 2. The top four diagnoses (urinary tract infection [UTI], sinusitis, upper respiratory infection [URI], and conjunctivitis) comprised 87% (877/1009) of all eVisits.

Because UTI and vaginitis diagnoses are limited to female patients in our eVisit menu, we reviewed demographics for a subset of eVisits with all eVisits for UTI and vaginitis removed. After removing eVisits for UTI and vaginitis, 565 eVisits remained. For this subset, 445 (79%) were females and the average age was 35.5 years. In this subset of eVisits, 75 (13%) were <18 years old.

Over half of patients, 583 (58%), received a prescription at the time of the eVisit. Prescribed medications were not limited to antibiotics, but included prescriptions for medications such as antivirals, cough suppressants, analgesics, intranasal steroids, and antihistamine eye drops. Frequency of type of prescription is shown in Table 3. Treatment rates with a prescription varied significantly between diagnosis type (p < 0.001 for difference between diagnosis categories) (Table 2).

Follow-up within 30 days of the eVisit by any method was seen in 340 patients (34%) with follow-up per diagnosis listed in Table 2. Two hundred twenty-seven (N = 227; 22%) patients were advised by the eVisit provider to have a follow-up visit to have their health concern addressed. Of these, 186 (82%) complied with these recommendations and did have subsequent follow-up within 30 days. Of the 41 (18%) who did not follow-up as advised, none were provided a prescription at the time of the eVisit. Of the remaining 782 (78%) patients who completed an eVisit without specific provider recommendations for follow-up, 154 (20%) had follow-up contact within the following 30 days. Of the 340 patients who had follow-up, 279 (82%) did so within the first 7 days of the eVisit. The mean time until follow-up was 1.2 days (interquartile range 0–0 days). Gender was not associated with follow-up care (34% of women had follow-up care versus 35% of men, p = 0.75). Types of follow-ups are shown in Table 4.

Over half of follow-up (59% [201/340]) was with a patient's primary care team (via phone call, portal message, or F2F primary care visit). Follow-up that occurred with F2F care (at the primary care clinic, Express Care retail clinic, and/or other specialty areas) comprised 74% (252/340) of follow-up care. Twenty-one patients (21/340 [6%]) had more than one follow-up contact type. Thirteen patients (13/340 [4%]) had F2F follow-up care in specialty areas such as gynecology, otolaryngology, and ophthalmology, in addition to one visit to an associated urgent care center.

Follow-up was more common in those who had not received a prescription at the time of the eVisit. Two hundred thirty-nine (239) of the 426 patients (56%) who did not receive a prescription at the time of the eVisit had follow-up within 30 days, while only 101 (17%) of the 583 patients who did receive a prescription at the time of the visit had follow-up within 30 days (p < 0.0001).

Fourteen patients received follow-up care in the emergency department (ED) with the following diagnosis at the eVisit represented: sinusitis (4), URI (4), and UTI (6), (p = 0.79 for diagnosis type). Eight of these patients had been told that they needed to seek additional evaluation with a F2F visit. Nine of the follow-up visits that occurred in the ED occurred after clinic hours when there was no other option for care available. There was one patient who was hospitalized and no deaths within 30 days of any patients who had completed eVisits for any cause.

Discussion

This 9-month review of eVisit utilization and follow-up revealed that the majority of eVisit requests were submitted during clinic hours. While one would speculate that eVisits would serve as an alternative to F2F care when such care is not available (i.e., weekends, holidays, and night hours), similar to a previous study at our institution, we noted that the majority of visits (70%) occurred during times when primary care offices were in operation. ¹⁴ This suggests that the convenience of an eVisit was a driving force for the patient's selection of this healthcare option and not simply availability of alternatives. Same-day appointments with a primary care provider are not always an option, and the risk of an undetermined wait time in the waiting room of a retail clinic or emergency department can be a deterrent to obtaining healthcare. As limited healthcare access is a national problem, ¹⁵ eVisit utilization likely will only increase and become an important adjunct to primary care for selected healthcare needs.

In our study, eVisits were primarily utilized by female patients. This is at least partly due to the fact that the conditions of UTI and yeast infection (which together comprised almost half of all eVisits) are limited to female patients only. However, even after controlling for female-specific illnesses, eVisits were still utilized far more commonly by female patients, a finding consistently noted in other eVisit studies. ^3,5,14,16 The majority of patients were of working age, reflecting a group that likely feels comfortable using technology and often have busy schedules. Pediatric patients account for over a quarter of retail clinic visits ¹⁷ ; however, this was not the case for eVisits in our study. Pediatric patients were likely underrepresented in our sample, as eVisits for sore throat were excluded as noted above. Of the 138 eVisits for sore throat, 36% were requested for patients under age 18 (compared with 7% of the remaining eVisits analyzed in our study). Other childhood illnesses (including ear infection and rash) that are commonly seen in retail clinics are not offered through our eVisit process because of the need to visualize such conditions for accurate diagnosis. Furthermore, at our institution, parents are unable to complete eVisits for children over 12 years of age due to limitations on parent's access to a minor patient's online portal; however, the child who is over 12 years of age may initiate their own eVisit.

Two-thirds of patients (66%) appeared to have had their concern resolved through the eVisit, as evidenced by not requiring any follow-up care, whether it was advised or not. Assuming the patients would have sought F2F care for their illnesses through another avenue, this represents significant cost savings for the consumer and insurer. eVisits on average cost less than $50 dollars, ^2,6 this is less expensive than a retail clinic or office visit, which in the instance of a UTI, may come with the additional cost of laboratory testing. A similar institution found a savings of $82–$142 when comparing eVisits with a traditional office visit and $159–$469 compared with an emergency department visit. ² Some healthcare institutions treat minor acute illnesses through phone or electronic messaging without a charge to the patient. The eVisit allows for this cost to the institution to be captured. It has been argued, however, the convenience of eVisits might lead to increased utilization, ¹⁴ as has been suggested in a study on retail clinics. ¹⁸ Patients may not have sought a F2F visit had the eVisit not been available. The fact that 18% of patients in our study who were advised to follow-up had no identified follow-up within 30 days could be supportive of this.

Overall, approximately one-third (34%) of patients had an additional contact with the healthcare system within 30 days of their eVisit for the same health concern, with the average time to follow-up just 1.2 days, with most following up within 24 h. When looking only at those not instructed to follow-up, this number is unchanged. The short time to follow-up suggests that patients are not worsening, but instead are seeking a different diagnosis or treatment. All phone calls and portal messages are documented in the patient's EMR. The majority of follow-up (74%) was through a F2F encounter. More than half of the patients who followed up (55%) were instructed by the eVisit provider to do so because it was felt a F2F examination was needed. When we exclude those advised to follow-up by the eVisit provider, the follow-up rate drops to 20%. This is similar to the follow-up rates found by Albert et al. (16.9%), although their study differed from ours in the types of diagnoses evaluated, a shorter follow-up time period (7 days), and our inclusion of any type of follow-up not limited to just F2F. ¹⁶

Having a follow-up contact was significantly associated with not receiving a prescription. Seventy percent of those who followed up did not receive a prescription at their eVisit. In the instance of URI, there were high follow-up rates and low prescription rates, while the opposite was true for sinusitis. This may, however, simply reflect provider bias with choosing the diagnosis of URI when the provider felt a prescription was not necessary versus diagnosing sinusitis when patient symptoms suggested treatment for a bacterial sinusitis. This may be an area for education—to the patient regarding prescribing expectations with certain symptoms and to the provider for management of acute sinusitis without antibiotics.

Overall, we found only a small number of emergency department visits associated with eVisits. Only one patient was hospitalized within 30 days. The patient sought treatment for UTI, and the eVisit provider directed the patient to the emergency department for severe symptoms. There were no deaths of any patients who had completed eVisits within the past 30 days. It should be noted that as only empaneled patients were included in the study, the providers had full access to the patient's EMR and were not reliant solely on data provided by the patient during the eVisit. This allowed for a thorough review of the patient's health history, including previous visits for this condition, potential medication interactions, and complicating comorbidities. This low rate of emergency department visits and hospitalization is similar to what was found by North et al. when both eVisits and secure electronic provider–patient messaging were studied. ¹⁹

Limitations include that there was no comparison of follow-up rates with a similarly matched cohort who presented for F2F care. Recent research is emerging to support that virtual care for minor upper respiratory illness may actually have better outcomes, in terms of lower follow-up care, than F2F visits for matched diagnoses. ⁸ Our study evaluated several minor acute illness diagnoses and evaluated eVisits that did not comprise virtual F2F contact with the patients. However, comparison with a F2F group would be important to see if the findings of virtual visits as a positive prognostic factor could be repeated and is an important area for future research. Another limitation to this study is that it was performed at a single institution. Our study would not have captured any follow-up care occurring at outside healthcare settings, although as only empaneled patients were included in our study, we anticipate this number would be low. Mayo Clinic is located in Olmsted County, MN, where there is only one other healthcare system in the county. Another potential limitation is that data were collected in the first year of eVisits using algorithmic questions. Anecdotally, a number of eVisit providers were uncomfortable with the idea of treating patients without an examination. This may have led to a greater number of patients being referred for F2F visits especially early on as providers grew more comfortable with providing care through eVisits.

Conclusions

eVisits are a method of healthcare delivery that will likely increase in utilization in the future. Our study supports that the majority of time that a patient seeks an eVisit for a minor acute illness the episode of care can be completed without any further interaction with the healthcare system.