Changes in Diagnostic and Demographic Characteristics of Patients Seeking Mental Health Care During the Early COVID-19 Pandemic in a Large, Community-Based Health Care System
Objective: The early COVID-19 pandemic resulted in great psychosocial disruption and stress, raising speculation that psychiatric disorders may worsen. This study aimed to identify patients vulnerable to worsening mental health during the COVID-19 pandemic.
Methods: This retrospective observational study used electronic health records from March 9 to May 31 in 2019 (n = 94,720) and 2020 (n = 94,589) in a large, community-based health care system. Percent change analysis compared variables standardized to the average patient population for the respective time periods.
Results: Compared to 2019, psychiatric visits increased significantly (P < .0001) in 2020, with the majority being telephone/video-based (+264%). Psychiatric care volume increased overall (7%), with the greatest increases in addiction (+42%), behavioral health in primary care (+17%), and adult psychiatry (+5%) clinics. While patients seeking care with preexisting psychiatric diagnoses were mainly stable (−2%), new patients declined (−42%). Visits for substance use (+51%), adjustment (+15%), anxiety (+12%), bipolar (+9%), and psychotic (+6%) disorder diagnoses, and for patients aged 18–25 years (+4%) and 26–39 years (+4%), increased. Child/adolescent and older adult patient visits decreased (−22.7% and −5.5%, respectively), and fewer patients identifying as White (−3.8%) or male (−5.0) or with depression (−3%) or disorders of childhood (−2%) sought care.
Conclusions: The early COVID-19 pandemic was associated with dramatic changes in psychiatric care facilitated by a rapid telehealth care transition. Patient volume, demographic, and diagnostic changes may reflect comfort with telehealth or navigating the psychiatric care system. These data can inform health system resource management and guide future work examining how care delivery changes impact psychiatric care quality and access.
J Clin Psychiatry 2021;82(2):20m13685
To cite: Ridout KK, Alavi M, Ridout SJ, et al. Changes in diagnostic and demographic characteristics of patients seeking mental health care during the early COVID-19 pandemic in a large, community-based healthcare system. J Clin Psychiatry. 2021;82(2):20m13685.
To share: https://doi.org/10.4088/JCP.20m13685
© Copyright 2021 Physicians Postgraduate Press, Inc.
aThe Permanente Medical Group, Kaiser Permanente Northern California, Oakland, California
bDivision of Research, Kaiser Permanente Northern California, Oakland, California
cKaiser Foundation Hospitals, Kaiser Permanente Northern California, Oakland, California
*Corresponding author: Kathryn K. Ridout, MD, PhD, 401 Bicentennial Way. Santa Rosa, CA 95403 (Kathryn.Erickson-Ridout@KP.org).
The coronavirus disease 2019 (COVID-19) pandemic has dramatically changed psychosocial environments, impacting the workplace, social interactions, finances, and daily routines worldwide. Most individuals report stress and fear related to these changes and to COVID-19 exposure and infection.1,2 Such stress and dramatic changes in daily functioning are large risk factors for new psychiatric symptoms or worsening preexisting psychiatric conditions, including depression, anxiety, posttraumatic stress, bipolar disorder, and substance use.3–6 While experts speculate that more people will seek psychiatric treatment during the COVID-19 pandemic,7–9 the true impact of COVID-19 on individuals’ psychiatric service utilization or conditions is unknown.
The current study examined changes in psychiatric care utilization and patient-level visit characteristics coinciding with orders by local and state authorities in northern California, United States, to shelter-in-place, which required individuals to remain at home except for activities deemed essential. We hypothesized that there would be more psychiatric service utilization compared to the same time period the year prior to the COVID-19 outbreak and that changes in patient-level psychiatric diagnoses and demographics would be seen.
Study Setting and Design
This retrospective, observational study was conducted in Kaiser Permanente Northern California (KPNC), a large community, primarily employer-based health system in the United States serving more than 4 million patients (an estimated 30% of the regional population). Data for 165,696 patient contacts among 94,720 unique patients in 2019 and 181,015 contacts among 94,589 patients in 2020 were included. As the Bay Area/California shelter-in-place orders commenced on March 16/19, 2020, and continued past May, the study inclusion criteria were patients seeking outpatient psychiatric care from March 9, 2020, to May 31, 2020, and the corresponding dates in 2019. Prior to the study time period, KPNC had already established telephone and video visits as options for accessing psychiatric care.10,11 All outpatient in-person and telehealth (video or telephone) encounters with psychiatric care providers in adult psychiatry, addiction, child and adolescent psychiatry, and behavioral health in primary care clinics were included. KPNC patients are highly representative of the ethnic and socioeconomic diversity of the surrounding and statewide population.12 The Research Determination Committee for the KPNC region determined that the project does not meet the regulatory definition of research involving human subjects per 45 CFR 46. 102(d).
Data regarding patient service utilization (contact type [phone, video, or in-person contact], provider type, and clinic type) and patient-level characteristics (visit-associated psychiatric diagnosis, patient history of previous psychiatric diagnosis, age, race/ethnicity, and reported gender) were extracted from electronic health record (EHR) data. Provider types were split into 3 categories: medication providers (board-certified or eligible physicians with a medical or doctor of osteopathy degree and nurse practitioners), therapists (psychologists, licensed family therapists, and licensed clinical social workers), and therapists specific to the addiction treatment department. Primary visit-associated psychiatric diagnoses, which were assigned by the treating psychiatrists, nurse practitioners, or therapists, were categorized according to International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) diagnostic codes (anxiety disorder: F40–43.12; depressive disorder [not included in bipolar or adjustment disorders]: F32–39; bipolar disorder: F30–31; psychotic disorder: F20–29; intellectual disabilities, developmental disorders, and disorders seen in childhood: F70–98; substance use disorder: F10–19; adjustment disorder: F43.2–43.9; and other disorders: all other codes [with diagnoses individually representing < 1% of patient visits]). To examine patient-level characteristics during in-person and telehealth visits, we examined a subset of patients (n = 82,864 in 2019 and n = 82,891 in 2020) with continuous KPNC enrollment during the year prior, which allowed for more complete ascertainment of these variables.
To assess differences in symptom severity between the two time periods, we tabulated any emergency or inpatient encounter with a primary admitting diagnosis for mental health. We included all ICD-10 diagnosis codes for mental health (including suicidal ideation) while excluding those for tobacco usage (See Supplementary Table 1 for complete list of ICD-10 codes). If the patient experienced multiple outcomes in the 3-month follow-up period, we gave precedence to the suicide diagnosis if that occurred in addition to any other mental health diagnosis.
Primary Data Analysis
We present continuous variables as means with 95% confidence intervals (CIs). We used the 12-week patient service utilization data from March 9 through May 31 from 2019 and 2020 to calculate psychiatric visit incidence ratios standardized by the average KPNC population for the respective time periods. A 2-sided P < .05 was considered statistically significant. Patient service utilization data were further used to graph proportion of visits (percent per 100,000 persons) for in-person and telehealth encounters for the study time periods. All analyses were conducted with SAS (version 9.31; SAS Institute, Inc; Cary, North Carolina).
Transition to Telehealth Care
Against the backdrop of an established telehealth infrastructure, telehealth care (telephone and video visits) increased by 264.4% (95% CI, 262.5% to 266.3%) during the study timeframe in 2020 compared to 2019, while in-person visits decreased by 92.1% (95% CI, −92.7% to –91.5%; Table 1). This change was concurrent with local/state shelter-in-place orders.
Psychiatric Service Utilization During the Early COVID-19 Pandemic
There was a slight decline in the number of unique patients in psychiatry treatment clinics during the COVID-19 pandemic period (−2.4%; 95% CI, −3.3% to −1.6%). However, there was an overall 6.7% increase (95% CI, 6.0% to 7.4%) in psychiatric service utilization during the 12 weeks after shelter-in-place orders compared to the same time period in 2019. Of those patients, the greatest increase was for visits in addiction treatment clinics (+42.0%; 95% CI, 39.4% to 44.6%) and with behavioral health providers in primary care (+16.5%; 95% CI, 11.9% to 21.1%) who can see/treat any psychiatric diagnosis, followed by adult psychiatry clinics (+4.7%; 95% CI, 4.0% to 5.5%), compared to 2019 levels. In contrast, child and adolescent psychiatry saw a 22.7% decrease in patient service utilization (95% CI, −25.6% to −19.8%). Patient contacts with addiction treatment counselors increased by 78.7% (95% CI, 71.9% to 85.4%), followed by therapy providers (+7.4%; 95% CI, 6.3% to 8.5%) and medication providers (+4.6%; 95% CI, 3.7% to 5.5%) in the 2020 versus the 2019 time periods, most of which was telehealth.
Visit-Level and Patient-Level Characteristics
We next examined visit-associated psychiatric diagnoses (in-person and telehealth visits) and patient demographics for KPNC members continuously enrolled during the year prior (Table 2). Relative to the 2019 period, there was an overall decrease in unique patients (−2.3%; 95% CI, −3.2% to −1.3%), although there was an increase in patient visits (+6.8%; 95% CI, 6.1% to 7.5%; Figure 1). This increase coincided with shelter-in-place orders, with overall visit incidence rates increasing to 300.8 encounters per 100,000 (95% CI, 299.4 to 302.3) in 2020 versus 282.1 per 100,000 (95% CI, 280.3 to 283.2) in 2019 (P < .0001). Although the median number of visits was 1 in both time periods, the mean number of visits significantly increased (1.90 versus 1.75, P < .0001).
During the early COVID-19 pandemic, 89% of patients had a psychiatric primary diagnosis as defined by ICD-10 codes, with visits related to substance use disorders increasing by 51.1% (95% CI, 47.8% to 54.4%), followed by adjustment (+14.5%; 95% CI, 11.3% to 17.7%), anxiety (+11.9%; 95% CI, 10.3% to 13.5%), bipolar (+8.5%; 95% CI, 5.8% to 11.2%), and psychotic (+6.4%; 95% CI, 2.8% to 9.9%) disorders, compared to the same time period in 2019. There were declines in the number of visits for depressive disorders (−3.1%; 95% CI, −4.4% to −1.8%) and intellectual disabilities/developmental disorders/disorders seen in childhood (−2.4%; 95% CI, −4.6% to −0.8%). Other primary diagnostic code categories included family/lifestyle health (−9.6%; 95% CI, −13.0% to −6.2%) and other medical diagnoses (+23%; 95% CI, 20.4% to 26.5%), which very likely reflects the increase in patient care by behavioral medicine in primary care clinics. Of the unique patients with a visit-associated psychiatric diagnosis, there was a 42.3% decrease in “new” patients (those with no psychiatric diagnosis ever documented in the EHR; 95% CI, −47.9% to −36.7%); “existing” patients who had a prior diagnosis sought care in similar numbers (1.6% decrease when adjusted by population; 95% CI, −2.6% to −0.6%). Visits increased by 4.3% (95% CI, 1.6% to 7.1%) and 3.8% (95% CI, 1.8% to 5.9%), respectively, among patients aged 18–25 and 26–39 years, but decreased among all other age groups, especially children aged 0–3 (−23.1%; 95% CI, −37.3% to −8.9%) and 4–12 years (−16.6%; 95% CI, −19.9% to −13.3%), adolescents (−5.4%; 95% CI, −8.2% to −2.7%), and older adults (−5.5%; 95% CI, −8.1% to −3.0%) relative to 2019 levels. Slightly fewer patients identifying as White (−3.8%; 95% CI, −5.1% to −2.6%) or male (−5.0%; 95% CI, −6.5% to −3.5%) sought care.
Symptom Severity and Suicidality
To understand differences in symptom severity between the COVID-19 time period and the year prior, data regarding emergency department and inpatient service utilization for a mental health diagnosis in the 3 months following the outpatient encounter were examined. Further, we analyzed the percentage of these visits that were for suicidal thoughts or behaviors. Compared to 2019, there was a 20.8% increase in emergency department visits and a 30.5% increase in inpatient hospitalizations (Table 3; 95% CI, 16.8% to 24.8% and 25.0% to 36.0%, respectively). Of these visits, the percentage for a chief complaint of suicidal thoughts or behaviors increased by 33.5% in the emergency department and 93.5% in the inpatient setting (95% CI, 26.5% to 40.6% and 60.6% to 126.3%, respectively).
The early COVID-19 pandemic shelter-in-place restrictions corresponded with dramatic shifts in psychiatric service utilization and care delivery within a large US health system. Telehealth care, formerly comprising a minority of visits, became the dominant mode of delivering care compared to the same time period in 2019. During this transition, the total volume of psychiatry visits increased, the total number of unique patients decreased, and patients with a preexisting psychiatric diagnosis were more represented relative to the comparison period. Increased patient service utilization may reflect an overall increased need for services, higher symptom load, increased care access through telehealth, or adaptation by the health system to maintain contact with patients with documented psychiatric disorders. The slight decrease in patients across the two time periods—despite unprecedented psychosocial disruptions imposed by the outbreak—suggests potential difficulties in reaching some individuals needing services.
These changes in care very likely relate to the COVID-19 pandemic. Outbreaks can result in significant mental stress and psychosocial disruption, impacting mental health. Worsening mental health was reported with the Ebola,13 influenza H1N1,14 and SARS outbreaks.15 Survey data suggest individuals report psychological distress, anxiety, and psychiatric concerns during the COVID-19 pandemic.1,2,16 Our data suggest that overall service utilization with psychiatric providers may increase when care is accessible to patients. The transition to primarily telehealth may facilitate patient care access by reducing barriers such as travel time, requesting time off work, or finding child care.
Subgroup analyses suggested that children, adolescents, older adults, and individuals with new psychiatric needs may have difficulties engaging in care during the pandemic. The overall drop in unique patient numbers coincided with a higher frequency of visits per patient, which may reflect a growth in treatment need or greater feasibility associated with telehealth visits for providers and patients in this insured population. Changes in child/adolescent and older adult engagement may relate to technology difficulties or barriers inherent to a telehealth platform or reflect reliance on caregivers to participate in visits.
Consistent with the increased utilization of addiction and behavioral health/primary care services, visit-associated diagnoses of substance use, adjustment, and anxiety disorders also increased. A decrease in visits related to depressive disorders contrasts with other reports of increased depressive symptoms associated with the pandemic,16 but may reflect increased endorsement of pandemic-related anxiety symptoms among patients with depressive disorders given that we captured only the primary visit diagnosis.
The nature and increased utilization of psychiatric care found in this study align with recent data1,2,16,17 reporting increases in anxiety symptoms and adjustment difficulties during the COVID-19 pandemic time period worldwide. Increases in disordered substance use have been hypothesized, but published data are lacking.18 Research on COVID-19 pandemic–related psychiatric service use has been sparse, mainly focusing on widespread transitions to telehealth care.17 The current study demonstrates that telehealth care can be used to meet increasing psychiatric care demand when in-person care is not feasible. Telehealth psychiatric care may be preferable for some patient populations (eg, established patients, younger adults) and less preferable for others (eg, new patients, children, adolescents, older adults).19 Previous work20 suggests that telehealth psychiatric care is as effective as in-person care. Future studies should examine if transitions to telehealth psychiatric care during the COVID-19 pandemic era have affected care quality or outcomes.
A limitation of the current study is that health care utilization changes among certain patient categories could reflect a range of factors including changes in individuals’ symptoms, differential access to care or care delivery, clinic outreach, or societal changes. For example, the slight decrease in visits among older adults may suggest challenges in accessing care due to less comfort with telehealth technology, but could also reflect lower endorsement of COVID-19 pandemic–related worries in this demographic group compared to younger individuals.1 Studies using patient-reported measures will be needed to understand potential disparities in telehealth care access across demographic or diagnostic categories. This population is an all-insured population and so may not be representative of uninsured populations. The study also takes place in the dynamic context of health care changing to strengthen telehealth. Our results may be influenced by unmeasured confounders that impact variables used to measure service utilization, such as changes in visit coding, care delivery (multiple visits/outreach), and appointment reminders that may vary by medical facility. These changes are difficult to quantify during the early phase of the pandemic as health systems seek to develop systematic approaches. While the majority of providers switched from a primarily in-person format to telehealth with COVID-19, changes observed in this study may differ based on provider comfort with telehealth. However, these data still offer insights on the changing landscape for health care delivery systems.
As the COVID-19 pandemic containment measures evolve, individuals and health care systems will continue to adapt. A challenge highlighted by the current work is how to reach individuals with emerging psychiatric symptoms who lack prior contact with psychiatric services. The decreases we observed among new patients seeking psychiatric care echo decreases in health care–seeking behavior seen for other medical disorders during the COVID-19 outbreak crisis.21–23 New patients may face difficulties establishing care with previously unfamiliar psychiatric providers through telehealth communication. Health systems can address the new emergence and exacerbation of psychiatric conditions during the COVID-19 pandemic era through public awareness campaigns and by promoting psychiatric treatment as part of patients’ overall health. Treatment models such as collaborative care based in primary care, which uses population management strategies and multidisciplinary care teams to treat common psychiatric conditions including depression and anxiety,24 may prove useful in responding to these demands. Further studies are needed to examine psychiatric care needs as the COVID-19 pandemic evolves, if COVID-19 directly exacerbates psychiatric symptoms, and the impact of rapid and potentially lasting psychiatric care delivery changes on care quality and outcomes.
Submitted: September 10, 2020; accepted January 5, 2021.
Published online: March 3, 2021.
Potential conflicts of interest: The authors report no financial or other relationship relevant to the subject of this article.
Funding/support: This work received internal grant funding from Kaiser Permanente Northern California’s Division of Research. Dr K. Ridout’s time was supported by The Permanente Medical Group’s Physician Researcher Program.
Role of the sponsor: The supporters had no role in the design, analysis, interpretation, or publication of this study.
Supplementary material: Available at Psychiatrist.com.
- Understanding the impact of the COVID-19 pandemic on psychiatric care can inform resource management and care delivery during the pandemic.
- During the COVID-19 pandemic, patients with substance use, adjustment, anxiety, bipolar, and psychotic disorders, and those aged 18–39 years, may seek care.
- Decreased visits from children and adolescents, older adults, or patients identifying as White or male or with depression may reflect opportunities for outreach.
Buy this Article as a PDF