E-visits for isotretinoin management have no effect on interruptions in care.

To the Editor: Although isotretinoin is highly effective for acne treatment, it requires burdensome in-person monthly visits.1Mori W.S. Houston N. Moreau J.F. et al.Personal burden of isotretinoin therapy and willingness to pay for electronic follow-up visits.JAMA Dermatol. 2016; 152: 338-340Crossref PubMed Scopus (19) Google Scholar In fact, missed appointments are a commonly cited reason for early treatment termination.2Shah N. Smith E. Kirkorian A.Y. Evaluating the barriers to isotretinoin treatment for acne vulgaris in pediatric patients.J Am Acad Dermatol. 2020; 85: 1597-1599Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar In recent years, amplified by the COVID-19 pandemic, virtual care for isotretinoin has rapidly expanded.3Kane S. Admani S. COVID-19 pandemic leading to the accelerated development of a virtual health model for isotretinoin.J Dermatol Nurses Assoc. 2021; 13: 54-57Crossref PubMed Scopus (7) Google Scholar Studies show that clinicians feel comfortable adjusting doses and monitoring patients receiving isotretinoin using an asynchronous e-visit model.4Das S. Su M.Y. Kvedar J.C. Smith G.P. Asynchronous telemedicine for isotretinoin management: a direct care pilot.J Am Acad Dermatol. 2022; 86: 184-186Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar However, little work has explored how e-visits affect continuity of isotretinoin management. Isotretinoin e-visits were instituted at the University of Michigan in July 2019. In this model, patients are initiated to receive isotretinoin at an office or via a synchronous video visit. Then, clinicians determine whether the patient case is appropriate for subsequent e-visits. An e-visit is a patient-initiated, asynchronous store-and-forward submission via the electronic medical record patient portal. Patients submit a questionnaire indicating progress, side effects, and photographs of the involved areas. Clinicians respond asynchronously with recommendations and prescriptions. Given that isotretinoin in-person visits have historically been associated with missed appointments and disruptions in care, this study aimed to determine if isotretinoin e-visits could improve continuity by decreasing interruptions in care.2Shah N. Smith E. Kirkorian A.Y. Evaluating the barriers to isotretinoin treatment for acne vulgaris in pediatric patients.J Am Acad Dermatol. 2020; 85: 1597-1599Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar We reviewed records of all acne patients treated with isotretinoin by our department in 2 time periods: from July 2018 to June 2019 (office visits only) and October 2019 to March 2021 (office, video, and e-visits). Interruption rate, defined as >1 week without an isotretinoin prescription, was calculated for each study period and then stratified by the visit subtype and demographics. Length of interruption was categorized as ≤1 month or >1 month. A total of 3146 encounters for isotretinoin were observed; 569 interruptions occurred. The interruption rate was similar between July 2018 and June 2019 (278/1462, 19.0%) and between October 2019 and March 2021 (291/1684, 17.3%) (P = .21), which held true across most demographics (Table I). From October 2019 to March 2021, e-visits had a similar interruption rate (150/870, 17.2%) to both office (103/583, 17.7%) and video visits (38/231, 16.5%) (P = .92) that was consistent across most demographics (Table II). Most interruptions led to a ≤1 month delay (435/569, 76.4%) versus >1 month (134/569, 23.6%), which did not differ between time periods (P = .147) or visit subtypes (P = .227).Table IInterruptions in care between 2 time periods: July 2018 to July 2019 (office visits only) and October 2019 to March 2021 (office, video, and e-visits)Percent of interruptions (interruptions/total visits)July 2018 to July 2019October 2019 to March 2021P valueOverall19.0 (278/1462)17.3 (291/1684).21Sex Male18.7 (136/727)18.4 (137/743).90 Female19.3 (142/735)16.4 (154/941).12Age (y) <1817.9 (93/520)20.5 (114/557).28 ≥1819.6 (185/942)∗Represents a significant difference between interruption rates in that demographic subgroup.15.7 (177/1127)∗Represents a significant difference between interruption rates in that demographic subgroup..02Insurance type Private17.7 (209/1183)14.9 (181/1213).69 Medicaid24.1 (56/232)24.1 (102/423).99 Medicare33.3 (13/39)16.7 (8/48).07 Other Governmental insurance0 (0/8)NA—Race Asian18.9 (20/106)15.6 (28/179).48 Black24.4 (21/86)30.4 (38/125).34 Alaska native/American Indian22.2 (2/9)25.0 (2/8)1.00 White18.3 (217/1186)16.1 (205/1271).16 Other, not listed, or patient refused24.0 (18/75)17.8 (18/101).32Analysis was conducted using a 2-tailed χ2 test.P values < .05 were considered significant.∗ Represents a significant difference between interruption rates in that demographic subgroup. Open table in a new tab Table IIInterruption rate of different visits stratified by patient demographic data from October 2019 to March 2021 (office, video, and e-visits)Percent of interruptions (interruptions/total visits)DemographicsE-visitOffice visitVideo visitP valueAll17.2 (150/870)17.7 (103/583)16.5 (38/231).92Sex Male19.9 (69/346)17.0 (47/277)17.5 (21/120).61 Female15.5 (81/524)18.3 (56/306)15.3 (17/111).53Age (y) <1819.7 (47/238)20.3 (46/227)22.8 (21/92).82 ≥1816.3 (103/632)16.0 (57/356)12.2 (17/139).48Insurance type Private15.9 (104/653)14.5 (59/407)11.8 (18/153).41 Medicaid21.8 (43/197)27.2 (43/157)23.2 (16/69).47 Medicare15.0 (3/20)∗Represents a significant difference when compared to e-visit interruption rate in that demographic subgroup. P values < .05 were considered significant.5.3 (1/19)∗Represents a significant difference when compared to e-visit interruption rate in that demographic subgroup. P values < .05 were considered significant.44.4 (4/9)∗Represents a significant difference when compared to e-visit interruption rate in that demographic subgroup. P values < .05 were considered significant..03Race Asian17.9 (15/84)15.5 (11/71)8.3 (2/24).53 Black28.1 (16/57)31.4 (16/51)35.3 (6/17).835 Alaska Native/American Indian40 (2/5)0 (0/3)0 (0/0).21 White16.1 (110/684)16.5 (68/411)15.3 (27/176).94 Other, not listed, or patient refused17.5 (7/40)17.0 (8/47)21.4 (3/14).93Analysis was conducted using a 2-tailed χ2 test. Significant findings were followed with binary logistic regression analysis.∗ Represents a significant difference when compared to e-visit interruption rate in that demographic subgroup. P values < .05 were considered significant. Open table in a new tab Analysis was conducted using a 2-tailed χ2 test. P values < .05 were considered significant. Analysis was conducted using a 2-tailed χ2 test. Significant findings were followed with binary logistic regression analysis. Overall, we found no change in the interruption rate or duration after isotretinoin e-visits were instituted and between visit subtypes. Similarly, we saw no differences across minority racial groups or those with Medicaid insurance. This indicates noninferiority of the e-visit process regarding continuity of care, within an appropriately selected patient population. On further investigation, we discovered appointment reminders for e-visits were not consistently sent when compared with video and office visits. Our system has since been modified to ensure that reminders to complete e-visits are always sent (Supplementary Fig 1, available via Mendeley at https://data.mendeley.com/datasets/4pjs92pt22/1). Limitations of this study include its single-site nature and that results may not be generalizable. Also, it was a retrospective review and not controlled. Additionally, patients opted into the e-visit model, which may introduce bias. Although there are barriers to implementing isotretinoin e-visits (provider compensation, patient access to technology with photo capabilities, etc), continuity of care does not seem to be affected by e-visits. Further studies are needed to determine the necessary optimizations for isotretinoin e-visits to improve continuity of therapy. None disclosed.