Risk factors for loss of follow-up after asynchronous dermatology eConsult concerning for skin cancer

Asynchronous electronic consultations (e-consults) can be a useful tool for the screening of cutaneous lesions, but may offer a malpractice risk. We characterized factors affecting initial eConsult office follow-up in a cohort of patients with documented neoplasm of uncertain behavior. Patients with an ICD 10 code of neoplasm with uncertain behavior (D48.5) at The Ohio State University that received an E-consult order from May 2017 to May 2021 were queried. Information collected included patient demographics, status of follow-up in-office appointment, referral status, and health care utilization. In-office follow-up appointments were defined as completed, cancelled/no-show or no-contact. 667 patients with a diagnosis of D48.5 were identified as having completed an eConsult. 427 (64%) patients had a documented phone/electronic message notifying the patient of the results of the eConsult. Year of encounter (0.88 [0.79–0.97]) and number of previously completed ambulatory visits (0.86 [0.77–0.96]) were significantly associated with documentation of phone/electronic message in the univariate and multivariate model. 429 (84%) patients had a dermatology office follow-up encounter while 82 (16%) had no appointment scheduled. Language spoken, referral status and race were significant in the univariate model, though race was the only significant variable in the multivariate model (P < 0.003). Asynchronous electronic consults to assess possible cutaneous neoplasms is an important tool for population screening of skin cancer. Dermatologists and health systems implementing an eConsult model for screening purposes should be aware of risk factors for loss of follow-up. Additional systems need to be implemented to ensure minorities and non-native English speakers are obtaining adequate dermatologic care.


Introduction
Asynchronous electronic consultations (eConsults) have been associated with improved metrics in the screening for skin cancer [1]. However, they may offer malpractice risk in documenting potential skin cancers in patients who may not follow-up for definitive diagnosis and treatment [2]. In this study, we sought to characterize risk factors associated with not completing office follow-ups in a cohort of patients after an eConsult documented suspicion for skin cancer.

Methods
Patients at The Ohio State University that received an eConsult to dermatology from May 2017 to May 2021 and were given a coded diagnosis of 'neoplasm of uncertain behavior' (D48.5) were queried. Information collected included patient demographics, preferred language, zip code affluence, year, eConsult initiating provider, and health care utilization. These were used as independent variables in forward stepwise models using final patient disposition as the dependent variable. Office follow-up appointments were defined as completed, patient cancelled/no-show, or as no confirmed contact.

Results
667 patients were identified as having a completed eConsult with the diagnosis of D48.5, of which 427 (64%) had a documented phone or electronic message notifying the patient of the results of the completed eConsult ( Table 1).
Year of encounter (0.88 [0.79-0.97]) and the number of previously completed ambulatory visits (0.86 [0.77-0.96]) were found to be significantly associated with documentation of phone/electronic message and remained significant in the multivariate model. Gender, affluence, referral status, race and patients' preferred language were not significantly associated with documented phone/electronic message. 429 (84%) patients had a dermatology office follow-up encounter while 82 (16%) had no appointment scheduled (  Table 1). There was no association found between clinic follow-up and affluence, year of encounter, or health care utilization.

Discussion
The ethical and malpractice concerns necessitate due diligence ensuring concerning neoplasms receive definitive diagnoses. As such, in 2018, despite the presence of a referral queue, schedulers began creating a specific note documenting patient contact attempts. This correlated with successfully improved documentation as noted in our study. Patients with more visits within our health system were also more likely to have documentation, likely due to more reliable telephone numbers and electronic messaging portal access. Most importantly, we can demonstrate that initial attempts of contact are not associated with race or patients' native language. Although race was ultimately the only significant association in the follow-up and cancel/no-show clinic models, many diverse ethnicities and backgrounds were represented and grouped into broad categories. We noted a lower likelihood of patients of Asian and other backgrounds completing a follow-up clinic visit as well as a higher likelihood of individuals identifying as Black to cancel/no show an office visit, which could be related to less cultural emphasis of skin cancer [3,4]. However, these groups are usually at a lower incidence of skin cancer [3], and they could note increased resolutions of lesions, an outcome that we do not have a method of confirming.
Dermatologists and healthcare systems implementing an eConsult model for screening purposes should be aware of risk factors for patients to not attend a recommended follow-up visit. We advocate for specific documentation and tracking methods to ensure the initial contact is made, although additional systems need to be implemented to ensure skin cancers do not go untreated, especially among patient racial minorities undergoing eConsults.
Author contributions SH and BK wrote the main manuscript text, prepared the figures, and performed data analysis. All authors reviewed the manuscript.
Funding None.

Conflict of interest
The authors declare no competing interests.