Kathryn E. Kemper

Kathryn E. Kemper

Auto-generated profile, based on publicly available information

Kathryn E. Kemper

Department of Epidemiology and Biostatistics, University of California, San Francisco, 550 16th St., Floor 2, San Francisco, CA, 94143, United States, 1 415-476-2300

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 March 2021 in International Journal of Gynecology & Obstetrics
Verified
Journal Article
Closed access
PDF file
KKKathryn E. Kemper
Kathryn E. Kemper
CMChristine J. McGrath
Christine J. McGrath
LELinda O. Eckert
Linda O. Eckert
JKJohn Kinuthia
John Kinuthia
BSBenson Singa
Benson Singa
ADAlison L. Drake
Alison L. Drake

22 August 2023 in JMIR Medical Informatics
Verified
Journal Article
Open access
SMShivani Mehta
Shivani Mehta
CLCourtney Rees Lyles
Courtney Rees Lyles
ARAnna Rubinsky
Anna Rubinsky
KKKathryn E Kemper
Kathryn E Kemper
JAJudith Auerbach
Judith Auerbach
WIWilliam Brown Iii
William Brown Iii
Background: Electronic health records (EHRs) have yet to fully capture social determinants of health (SDOH) due to challenges such as nonexistent or inconsistent data capture tools across clinics, lack of time, and the burden of extra steps for the clinician. However, patient clinical notes (unstructured data) may be a better source of patient-related SDOH information. Objective: It is unclear how accurately EHR data reflect patients’ lived experience of SDOH. The manual process of retrieving SDOH information from clinical notes is time-consuming and not feasible. We leveraged two high-throughput tools to identify SDOH mappings to structured and unstructured patient data: PatientExploreR and Electronic Medical Record Search Engine (EMERSE). Methods: We included adult patients (≥18 years of age) receiving primary care for their diabetes at the University of California, San Francisco (UCSF), from January 1, 2018, to December 31, 2019. We used expert raters to develop a corpus using SDOH in the compendium as a knowledge base as targets for the natural language processing (NLP) text string mapping to find string stems, roots, and syntactic similarities in the clinical notes of patients with diabetes. We applied advanced built-in EMERSE NLP query parsers implemented with JavaCC. Results: We included 4283 adult patients receiving primary care for diabetes at UCSF. Our study revealed that SDOH may be more significant in the lives of patients with diabetes than is evident from structured data recorded on EHRs. With the application of EMERSE NLP rules, we uncovered additional information from patient clinical notes on problems related to social connectionsisolation, employment, financial insecurity, housing insecurity, food insecurity, education, and stress. Conclusions: We discovered more patient information related to SDOH in unstructured data than in structured data. The application of this technique and further investment in similar user-friendly tools and infrastructure to extract SDOH information from unstructured data may help to identify the range of social conditions that influence patients’ disease experiences and inform clinical decision-making.

24 March 2025 in Journal of Medical Internet Research
Verified
Journal Article
Open access
JSJonathan J Shih
Jonathan J Shih
MKMagdalene Kuznia
Magdalene Kuznia
SNSarah Nouri
Sarah Nouri
ESElizabeth B Sherwin
Elizabeth B Sherwin
KKKathryn E Kemper
Kathryn E Kemper
EKElaine C Khoong
Elaine C Khoong
Background: The COVID-19 public health emergency catalyzed widespread adoption of both video- and audio-only telemedicine visits. This proliferation highlighted inequities in use by age, race and ethnicity, and preferred language. Few studies have investigated how differences in health system telemedicine implementation affected these inequities. Objective: This study aims to describe patients who used telemedicine during the public health emergency and identify predictors of telemedicine use across 2 health systems with different telemedicine implementations. Methods: This retrospective cohort study included adults with diabetes receiving primary care between July 2020 and March 2021 at 2 independent health systems in San Francisco, California. Participant sociodemographic characteristics, health information, and telemedicine utilization were acquired from electronic health records. The primary outcome was visit type (any audio or video telemedicine vs in-person only) during the study period. We used multivariable logistic regression to assess the association between visit type and key predictors associated with digital exclusion (age, race and ethnicity, preferred language, and neighborhood socioeconomic status), adjusting for baseline health. We included an interaction term to evaluate health system impact on each predictor and then stratified by health system (academic, which prioritized video-enabled visits, vs safety net, which prioritized audio-only visits). Results: Among 10,201 patients, we found higher odds of telemedicine use in the safety net system compared with the academic system (adjusted odds ratio [aOR] 2.94, 95% CI 2.48-3.48). Patients with younger age (18-34 years: aOR 2.55, 95% CI 1.63-3.97; 35-49 years: aOR 1.39, 95% CI 1.12-1.73 vs 75+ years) and Chinese-language preference (aOR 2.04, 95% CI 1.66-2.5 vs English) had higher odds of having a telemedicine visit. Non-Hispanic Asian (aOR 0.67, 95% CI 0.56-0.79), non-Hispanic Black (aOR 0.83, 95% CI 0.68-1), and Hispanic or Latine (aOR 0.76, 95% CI 0.61-0.95) patients had lower odds of having a telemedicine visit than non-Hispanic White patients. We found significant interactions between health system and age, race and ethnicity, and preferred language (P<.05). After stratifying by health system, several differences persisted in the academic system: non-Hispanic Asian (aOR 0.57, 95% CI 0.46-0.70) and Latine (aOR 0.67, 95% CI 0.50-0.91) patients had lower odds of a telemedicine visit, and younger age groups had higher odds (18-34 years: aOR 3.97, 95% CI 1.99-7.93; 35-49 years: aOR 1.86, 95% CI 1.36-2.56). In the safety net system, Chinese-speaking patients had higher odds of having a telemedicine visit (aOR 2.52, 95% CI 1.85-3.42). Conclusions: We found disparities in telemedicine utilization by age, race and ethnicity, and preferred language, primarily in the health system that used more video visits. While telemedicine expanded rapidly recently, certain populations remain at risk for digital exclusion. These findings suggest that system-level factors influence telemedicine adoption and implementation decisions impact accessibility for populations at risk for digital exclusion.

14 September 2023 in PLOS Global Public Health
Verified
Journal Article
Open access
PDF file
KKKathryn E. Kemper
Kathryn E. Kemper
OAOrvalho Augusto
Orvalho Augusto
SGStephen Gloyd
Stephen Gloyd
DADerick A. Akoku
Derick A. Akoku
GOGbossouna Ouattara
Gbossouna Ouattara
AKAhoua Koné
Ahoua Koné
Routine viral load (VL) monitoring is the standard of care in Côte d’Ivoire and allows for effective treatment guidance for people living with human immunodeficiency virus (HIV) to reach viral load suppression (VLS). For VL monitoring to be effective in reducing the impact of HIV, it must be provided in accordance with national guidance. This study aimed to evaluate VL testing, VLS rates and adherence to national guidance for VL testing using data collected from three national laboratories. We collected data on VL testing between 2015–2018 from OpenELIS (OE), an open-source electronic laboratory information system. We merged data by unique patient ID for patients (0–80 years old) who received multiple VL tests to calculate time between tests. We defined VLS as HIV RNA ≤1,000 copies/mL based on Côte d’Ivoire national and WHO guidance at the time of data collection. We used the Kaplan-Meier survival estimator to estimate time between ART (antiretroviral therapy) initiation and the first VL test, time between subsequent VL tests, and to estimate the proportion of people living with HIV (PLHIV) who were virally suppressed within 12 months of ART initiation. At the first documented VL test, 79.6% of patients were virally suppressed (95% CI: 78.9–80.3). Children under 15 were the least likely to be virally suppressed (55.2%, 95% CI: 51.5–58.8). The median time from ART initiation to the first VL sample collection for testing was 7.8 months (IQR:6.2–13.4). 72.4% of patients were virally suppressed within one year of treatment initiation (95% CI:71.5–73.3). Approximately 30% of patients received a second VL test during the 4-year study period. The median time between the first and second VL tests was 24.9 months (IQR: 4.7->40). Most PLHIV received their first VL test within the recommended 12 months of ART initiation but did not receive subsequent VL monitoring tests within the recommended time frame, reducing the benefits of VL monitoring. While VLS was fairly high, children were least likely to be virally suppressed. Our findings highlight the importance of regular VL monitoring after the first VL test, especially for children.

23 July 2024 in JMIR Preprints
Verified
Preprint
Open access
JSJonathan J Shih
Jonathan J Shih
MKMagdalene Kuznia
Magdalene Kuznia
SNSarah Nouri
Sarah Nouri
ESElizabeth B Sherwin
Elizabeth B Sherwin
KKKathryn E Kemper
Kathryn E Kemper
EKElaine C Khoong
Elaine C Khoong
BACKGROUND The COVID-19 public health emergency catalyzed widespread adoption of both video- and audio-only telemedicine visits. This proliferation highlighted inequities in use by age, race and ethnicity, and preferred language. Few studies have investigated how differences in health system telemedicine implementation affected these inequities. OBJECTIVE This study aims to describe patients who used telemedicine during the public health emergency and identify predictors of telemedicine use across 2 health systems with different telemedicine implementations. METHODS This retrospective cohort study included adults with diabetes receiving primary care between July 2020 and March 2021 at 2 independent health systems in San Francisco, California. Participant sociodemographic characteristics, health information, and telemedicine utilization were acquired from electronic health records. The primary outcome was visit type (any audio or video telemedicine vs in-person only) during the study period. We used multivariable logistic regression to assess the association between visit type and key predictors associated with digital exclusion (age, race and ethnicity, preferred language, and neighborhood socioeconomic status), adjusting for baseline health. We included an interaction term to evaluate health system impact on each predictor and then stratified by health system (academic, which prioritized video-enabled visits, vs safety net, which prioritized audio-only visits). RESULTS Among 10,201 patients, we found higher odds of telemedicine use in the safety net system compared with the academic system (adjusted odds ratio [aOR] 2.94, 95% CI 2.48-3.48). Patients with younger age (18-34 years: aOR 2.55, 95% CI 1.63-3.97; 35-49 years: aOR 1.39, 95% CI 1.12-1.73 vs 75+ years) and Chinese-language preference (aOR 2.04, 95% CI 1.66-2.5 vs English) had higher odds of having a telemedicine visit. Non-Hispanic Asian (aOR 0.67, 95% CI 0.56-0.79), non-Hispanic Black (aOR 0.83, 95% CI 0.68-1), and Hispanic or Latine (aOR 0.76, 95% CI 0.61-0.95) patients had lower odds of having a telemedicine visit than non-Hispanic White patients. We found significant interactions between health system and age, race and ethnicity, and preferred language (P<.05). After stratifying by health system, several differences persisted in the academic system: non-Hispanic Asian (aOR 0.57, 95% CI 0.46-0.70) and Latine (aOR 0.67, 95% CI 0.50-0.91) patients had lower odds of a telemedicine visit, and younger age groups had higher odds (18-34 years: aOR 3.97, 95% CI 1.99-7.93; 35-49 years: aOR 1.86, 95% CI 1.36-2.56). In the safety net system, Chinese-speaking patients had higher odds of having a telemedicine visit (aOR 2.52, 95% CI 1.85-3.42). CONCLUSIONS We found disparities in telemedicine utilization by age, race and ethnicity, and preferred language, primarily in the health system that used more video visits. While telemedicine expanded rapidly recently, certain populations remain at risk for digital exclusion. These findings suggest that system-level factors influence telemedicine adoption and implementation decisions impact accessibility for populations at risk for digital exclusion.

9 July 2026 in Journal of Medical Internet Research
Verified
Journal Article
Open access
DODebora L Oh
Debora L Oh
KKKathryn E Kemper-McIsaac
Kathryn E Kemper-McIsaac
DMDan Meltzer
Dan Meltzer
EBEric Brelsford
Eric Brelsford
KTKelsey Taylor
Kelsey Taylor
MDMindy C DeRouen
Mindy C DeRouen
Places have a significant impact on health. Thus, examining place-based structural and social determinants can help inform effective public health interventions. HealthAtlas.ucsf.edu provides a single online platform to explore multiple domains of place-based data. The initial California-only version of Health Atlas was launched in April 2020, and a national version was launched in October 2024. An artificial intelligence–assisted search function was integrated in January 2026. Health Atlas includes data on over 200 variables across 5 topical domains: demographics, socioeconomic, neighborhood, environment, and health and health care. The data were either (1) obtained from the American Community Survey, CDC PLACES, and other public sources or (2) obtained through collaboration with research partners. Users can visualize and aggregate data for 7 geographic levels: census tract, zip code tabulation area, county, congressional district, core-based statistical area, public use microdata area, and state. Most variables are available across all 50 states in the United States, the District of Columbia, and Puerto Rico. Users can explore Health Atlas to better understand the relationship between selected variables via maps, histograms, scatterplots, and summary plots. Users can select custom areas to suit specific needs. Customizable datasets can be downloaded for further use. Health Atlas provides freely available, user-friendly data and tools to support researchers, community organizations, government entities, and other public health professionals. We envision that Health Atlas will contribute to evidence-based, community-based initiatives; impactful health equity research; and effective public health programs.