Timothy Antipas Peter
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Timothy Antipas Peter has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Catherine Ann Staton
Duke Global Health Institute, Duke University, Durham, NC, USA
4 shared publications - Blandina Theophil Mmbaga
KCMC University, Moshi, Tanzania
3 shared publications - Anna Tupetz
Department of Emergency Medicine Duke University School of Medicine Durham North Carolina USA
2 shared publications - João Ricardo Nickenig Vissoci
Duke Global Health Institute, Duke University, Durham, NC, USA
2 shared publications - Alice Andongolile
Kilimanjaro Clinical Research Institute
1 shared publication - Anjni Joiner
Department of Emergency Medicine, Duke University
1 shared publication
Latest Publications
- Open Access
- 4 August 2026
- Published by Wiley in Health Science Reports
- Vol. 9 (8) , e72956
Predicting Outcomes of Traumatic Brain Injury Using Machine Learning Models Among Patients at Kilimanjaro Christian Medical Centre, Tanzania: A Registry‐Based Cohort Study
Background: Traumatic brain injury (TBI) remains a major global health burden, disproportionately affecting low‐ and middle‐income countries (LMICs) where access to neurocritical care is limited. Accurate and context‐appropriate prognostic models are crucial to guide early clinical decision‐making and optimize resource allocation in such settings. This study aims to develop and evaluate machine learning (ML) models for predicting TBI outcomes among adult patients using trauma registry data from Kilimanjaro Christian Medical Centre (KCMC), Tanzania. Methods: This retrospective cohort study utilized data from 4596 adult TBI patients recorded in the KCMC trauma registry between 2013 and 2024. The outcome was dichotomized Glasgow Outcome Scale (GOS): poor (1–3) versus good (4–5). Ten supervised ML algorithms, including Random Forest (RF), Decision Tree (DT), Logistic Regression, Support Vector Machine (SVM), and Artificial Neural Networks (ANN), were trained on 70% of the data after applying multiple imputation and synthetic minority Over‐sampling Technique (SMOTE) to address missingness and class imbalance. Hyperparameter tuning was performed using 10‐fold cross‐validation. Model performance was assessed on a 30% test set using area under the ROC curve (AUC), accuracy, sensitivity, specificity, and predictive values. Results: Among the 4596 patients, 26.2% had poor outcomes. The RF and DT models achieved the highest AUCs of 0.83 and 0.82, respectively. RF also showed the highest accuracy (0.78) and strong positive predictive value (PPV = 0.87), while DT had the highest sensitivity for poor outcomes (84.5%). Predictors of poor outcomes included TBI severity, pupil non‐reactivity, low oxygen saturation, lack of CT scan, alcohol use, and abnormal vital signs. Conclusion: ML models, particularly RF and DT, demonstrated strong predictive performance for TBI outcomes using routinely collected variables in a resource‐limited LMIC setting. Their interpretability and reliance on admission‐level data make them potential tools for real‐time triage and risk stratification. Future research should focus on external validation and integration into clinical decision‐support systems to support scaleup across similar settings.
- 4 August 2026
- Published by Wiley in Health Science Reports
- Vol. 9 (8) , e72956
Open AccessPredicting Outcomes of Traumatic Brain Injury Using Machine Learning Models Among Patients at Kilimanjaro Christian Medical Centre, Tanzania: A Registry‐Based Cohort Study
Background: Traumatic brain injury (TBI) remains a major global health burden, disproportionately affecting low‐ and middle‐income countries (LMICs) where access to neurocritical care is limited. Accurate and context‐appropriate prognostic models are crucial to guide early clinical decision‐making and optimize resource allocation in such settings. This study aims to develop and evaluate machine learning (ML) models for predicting TBI outcomes among adult patients using trauma registry data from Kilimanjaro Christian Medical Centre (KCMC), Tanzania. Methods: This retrospective cohort study utilized data from 4596 adult TBI patients recorded in the KCMC trauma registry between 2013 and 2024. The outcome was dichotomized Glasgow Outcome Scale (GOS): poor (1–3) versus good (4–5). Ten supervised ML algorithms, including Random Forest (RF), Decision Tree (DT), Logistic Regression, Support Vector Machine (SVM), and Artificial Neural Networks (ANN), were trained on 70% of the data after applying multiple imputation and synthetic minority Over‐sampling Technique (SMOTE) to address missingness and class imbalance. Hyperparameter tuning was performed using 10‐fold cross‐validation. Model performance was assessed on a 30% test set using area under the ROC curve (AUC), accuracy, sensitivity, specificity, and predictive values. Results: Among the 4596 patients, 26.2% had poor outcomes. The RF and DT models achieved the highest AUCs of 0.83 and 0.82, respectively. RF also showed the highest accuracy (0.78) and strong positive predictive value (PPV = 0.87), while DT had the highest sensitivity for poor outcomes (84.5%). Predictors of poor outcomes included TBI severity, pupil non‐reactivity, low oxygen saturation, lack of CT scan, alcohol use, and abnormal vital signs. Conclusion: ML models, particularly RF and DT, demonstrated strong predictive performance for TBI outcomes using routinely collected variables in a resource‐limited LMIC setting. Their interpretability and reliance on admission‐level data make them potential tools for real‐time triage and risk stratification. Future research should focus on external validation and integration into clinical decision‐support systems to support scaleup across similar settings.
- Open Access
- 21 November 2024
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 4 (11) , e0003942
Sex differences in alcohol use patterns and related harms: A mixed-methods, cross-sectional study of men and women in northern Tanzania
In northern Tanzania, alcohol use disorders (AUD) are under-diagnosed and under-treated, and current services are mostly limited to men in clinical settings despite significant alcohol-related harm in the community. The study objective was to identify sex differences in alcohol use and alcohol-related harms within and across community and clinical settings. This was a congruent triangulation mixed methods study consisting of focus group discussions (FGDs) and cross-sectional surveys. Quantitative analysis was conducted via Drinker Inventory of Consequences (DrInC) and Alcohol Use Disorders Identification Test (AUDIT) data from injury patients presenting for care at the Kilimanjaro Christian Medical Center Emergency Department and community participants. Differences in scores by sex were assessed using unpaired t-tests. K-means algorithms were run independently in both samples. Deductive thematic analysis was conducted on FGDs with community members, injury patients, and injury patient relatives. Differences in mean scores between sexes in the community and patient samples were statistically significant (p<0.05). Men showed higher AUDIT and DrInC mean scores in both samples. K-means separated the community and patient samples into two clusters, one with and one without harmful alcohol use. Of those indicating harmful alcohol use, the community cluster (n = 77, AUDIT = 14.29±9.22, DrInC = 22.67±6.80) was 27% women; the patient cluster (n = 57, AUDIT = 15.00±9.48, DrInC = 27.00±7.76) was 5% women. FGD transcripts revealed sex differences in four themes: alcohol initiation, consumption patterns, risk behaviors, and social stigma. This study identified important sex differences in the manifestation of AUD in northern Tanzania with respect to alcohol initiation, consumption patterns, risk behavior, and stigma. These findings indicate that women may need to be encouraged to seek injury care at the Emergency Department. Future research, prevention, and treatment efforts intended to reduce alcohol-related harms need to account for sex differences to optimize reach and effectiveness.
- 21 November 2024
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 4 (11) , e0003942
Open AccessSex differences in alcohol use patterns and related harms: A mixed-methods, cross-sectional study of men and women in northern Tanzania
In northern Tanzania, alcohol use disorders (AUD) are under-diagnosed and under-treated, and current services are mostly limited to men in clinical settings despite significant alcohol-related harm in the community. The study objective was to identify sex differences in alcohol use and alcohol-related harms within and across community and clinical settings. This was a congruent triangulation mixed methods study consisting of focus group discussions (FGDs) and cross-sectional surveys. Quantitative analysis was conducted via Drinker Inventory of Consequences (DrInC) and Alcohol Use Disorders Identification Test (AUDIT) data from injury patients presenting for care at the Kilimanjaro Christian Medical Center Emergency Department and community participants. Differences in scores by sex were assessed using unpaired t-tests. K-means algorithms were run independently in both samples. Deductive thematic analysis was conducted on FGDs with community members, injury patients, and injury patient relatives. Differences in mean scores between sexes in the community and patient samples were statistically significant (p<0.05). Men showed higher AUDIT and DrInC mean scores in both samples. K-means separated the community and patient samples into two clusters, one with and one without harmful alcohol use. Of those indicating harmful alcohol use, the community cluster (n = 77, AUDIT = 14.29±9.22, DrInC = 22.67±6.80) was 27% women; the patient cluster (n = 57, AUDIT = 15.00±9.48, DrInC = 27.00±7.76) was 5% women. FGD transcripts revealed sex differences in four themes: alcohol initiation, consumption patterns, risk behaviors, and social stigma. This study identified important sex differences in the manifestation of AUD in northern Tanzania with respect to alcohol initiation, consumption patterns, risk behavior, and stigma. These findings indicate that women may need to be encouraged to seek injury care at the Emergency Department. Future research, prevention, and treatment efforts intended to reduce alcohol-related harms need to account for sex differences to optimize reach and effectiveness.
- Open Access
- 26 February 2024
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 4 (2) , e0002946
Predictors of prior HIV testing and acceptance of a community-based HIV test offer among male bar patrons in northern Tanzania
Community-based HIV testing offers an alternative approach to encourage HIV testing among men in sub-Saharan Africa. In this study, we evaluated a community-based HIV testing strategy targeting male bar patrons in northern Tanzania to assess factors predictive of prior HIV testing and factors predictive of accepting a real-time HIV test offer. Participants completed a detailed survey and were offered HIV testing upon survey completion. Poisson regression was used to identify prevalence ratios for the association between potential predictors and prior HIV testing or real-time testing uptake. Of 359 participants analyzed, the median age was 41 (range 19–82) years, 257 (71.6%) reported a previous HIV test, and 321 (89.4%) accepted the real-time testing offer. Factors associated with previous testing for HIV (adjusted prevalence ratio [aPR], 95% CI) were wealth scores in the upper-middle quartile (1.25, 1.03–1.52) or upper quartile (1.35, 1.12–1.62) and HIV knowledge (1.04, 1.01–1.07). Factors that predicted real-time testing uptake were lower scores on the Gender-Equitable Men scale (0.99, 0.98–0.99), never testing for HIV (1.16, 1.03–1.31), and testing for HIV > 12 months prior (1.18, 1.06–1.31). We show that individual-level factors that influence the testing-seeking behaviors of men are not likely to impact their acceptance of an HIV offer.
- 26 February 2024
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 4 (2) , e0002946
Open AccessPredictors of prior HIV testing and acceptance of a community-based HIV test offer among male bar patrons in northern Tanzania
Community-based HIV testing offers an alternative approach to encourage HIV testing among men in sub-Saharan Africa. In this study, we evaluated a community-based HIV testing strategy targeting male bar patrons in northern Tanzania to assess factors predictive of prior HIV testing and factors predictive of accepting a real-time HIV test offer. Participants completed a detailed survey and were offered HIV testing upon survey completion. Poisson regression was used to identify prevalence ratios for the association between potential predictors and prior HIV testing or real-time testing uptake. Of 359 participants analyzed, the median age was 41 (range 19–82) years, 257 (71.6%) reported a previous HIV test, and 321 (89.4%) accepted the real-time testing offer. Factors associated with previous testing for HIV (adjusted prevalence ratio [aPR], 95% CI) were wealth scores in the upper-middle quartile (1.25, 1.03–1.52) or upper quartile (1.35, 1.12–1.62) and HIV knowledge (1.04, 1.01–1.07). Factors that predicted real-time testing uptake were lower scores on the Gender-Equitable Men scale (0.99, 0.98–0.99), never testing for HIV (1.16, 1.03–1.31), and testing for HIV > 12 months prior (1.18, 1.06–1.31). We show that individual-level factors that influence the testing-seeking behaviors of men are not likely to impact their acceptance of an HIV offer.
- Open Access
- 13 June 2022
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 2 (6) , e0000277
Barriers to accessing follow up care in post-hospitalized trauma patients in Moshi, Tanzania: A mixed methods study
Disproportionately high injury rates in Sub-Saharan Africa combined with limited access to care in both the acute injury phase and for injury patients requiring continued care after hospital discharge remains a challenge. We aimed to characterize barriers to transportation and access to care in a cohort of post-hospitalized injury patients in Moshi, Tanzania. This was a mixed-methods study of a prospective cohort of trauma registry patients presenting to Kilimanjaro Christian Medical Center between August 2018 and January 2020. We conducted standardized patient/family surveys and in-depth interviews at a 2-week follow up visit after hospital discharge, and focus groups with healthcare providers. Quantitative results were analyzed using descriptive statistics and multivariable logistic regression using R statistical software. Qualitative results were analyzed using thematic analysis through an iterative process using NVivo software. A total of 1,365 patients were enrolled in the trauma registry, with 169 patients followed up at 2 weeks. Over half of patients at follow-up, 101 (59.8%), reported challenges in traveling. The majority of patients were male (80.3%). Difficulty in traveling since injury was associated with female gender (aOR 5.85 [95% CI 1.20–33.59]) and a need for non-family members escorts for travel (aOR 7.10 [95% CI 1.43–41.66]). Those who reported assault or fall as the mechanism of injury as compared to road traffic injury and had health insurance were less likely to report challenges in traveling (aOR 0.19 [95% CI 0.03–0.90]), 0.11 [95% CI 0.01–0.61], 0.14 [95% 0.02–0.80]). Transportation barriers that emerged from qualitative data included inability to use regular means of transportation, financial challenges, physical barriers, rigid compliance to physician orders, access to healthcare, and social support barriers. Our findings demonstrate several areas to address transportation barriers for post-injury patients in Tanzania. Educational interventions such as clarification of doctors’ orders of strict bedrest, provision of vouchers to support financial challenges and alternate means of transportation given physical barriers and reliance on social support may address some of these barriers.
- 13 June 2022
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 2 (6) , e0000277
Open AccessBarriers to accessing follow up care in post-hospitalized trauma patients in Moshi, Tanzania: A mixed methods study
Disproportionately high injury rates in Sub-Saharan Africa combined with limited access to care in both the acute injury phase and for injury patients requiring continued care after hospital discharge remains a challenge. We aimed to characterize barriers to transportation and access to care in a cohort of post-hospitalized injury patients in Moshi, Tanzania. This was a mixed-methods study of a prospective cohort of trauma registry patients presenting to Kilimanjaro Christian Medical Center between August 2018 and January 2020. We conducted standardized patient/family surveys and in-depth interviews at a 2-week follow up visit after hospital discharge, and focus groups with healthcare providers. Quantitative results were analyzed using descriptive statistics and multivariable logistic regression using R statistical software. Qualitative results were analyzed using thematic analysis through an iterative process using NVivo software. A total of 1,365 patients were enrolled in the trauma registry, with 169 patients followed up at 2 weeks. Over half of patients at follow-up, 101 (59.8%), reported challenges in traveling. The majority of patients were male (80.3%). Difficulty in traveling since injury was associated with female gender (aOR 5.85 [95% CI 1.20–33.59]) and a need for non-family members escorts for travel (aOR 7.10 [95% CI 1.43–41.66]). Those who reported assault or fall as the mechanism of injury as compared to road traffic injury and had health insurance were less likely to report challenges in traveling (aOR 0.19 [95% CI 0.03–0.90]), 0.11 [95% CI 0.01–0.61], 0.14 [95% 0.02–0.80]). Transportation barriers that emerged from qualitative data included inability to use regular means of transportation, financial challenges, physical barriers, rigid compliance to physician orders, access to healthcare, and social support barriers. Our findings demonstrate several areas to address transportation barriers for post-injury patients in Tanzania. Educational interventions such as clarification of doctors’ orders of strict bedrest, provision of vouchers to support financial challenges and alternate means of transportation given physical barriers and reliance on social support may address some of these barriers.
- Open Access
- 3 June 2022
- Published by MDPI AG in Trauma Care
- Vol. 2 (2) , 341-358
“I Don’t Do Anything; I’m Just Being Taken Care Of”: Experiences of Patients and Their Caregivers Transitioning Back into the Community Following Traumatic Injury in Northern Tanzania
After discharge from the hospital for traumatic injury, patients and their caregivers face a period of increased vulnerability. This adjustment phase is poorly characterized, especially in low- and middle-income countries. We explored the experiences of patients and their caregivers in Northern Tanzania after hospitalization for a traumatic injury. Patients who received care for traumatic injury at the Kilimanjaro Christian Medical Center and their caregivers were selected as part of a convenience sample from January 2019 to December 2019. Analysts developed a codebook; content and analytic memos were subsequently created. We then applied the biopsychosocial model to further characterize our findings. Participants included 26 patients and 11 caregivers. Patients were mostly middle-aged (mean age 37.7) males (80.8%), residing in urban settings (57.7%), injured in road traffic accidents (65.4%), and who required surgery (69.2%). Most caregivers were female. Seven major themes arose: pain, decreased physical functioning, poor emotional health, lack of support, challenges with daily activities, financial strain, and obstacles to accessing healthcare. This study describes some of the difficulties transitioning back into the community after hospitalization for traumatic injury. Our work demonstrates the importance of mixed methods approaches in characterizing and addressing transitions of care challenges.
- 3 June 2022
- Published by MDPI AG in Trauma Care
- Vol. 2 (2) , 341-358
Open Access“I Don’t Do Anything; I’m Just Being Taken Care Of”: Experiences of Patients and Their Caregivers Transitioning Back into the Community Following Traumatic Injury in Northern Tanzania
After discharge from the hospital for traumatic injury, patients and their caregivers face a period of increased vulnerability. This adjustment phase is poorly characterized, especially in low- and middle-income countries. We explored the experiences of patients and their caregivers in Northern Tanzania after hospitalization for a traumatic injury. Patients who received care for traumatic injury at the Kilimanjaro Christian Medical Center and their caregivers were selected as part of a convenience sample from January 2019 to December 2019. Analysts developed a codebook; content and analytic memos were subsequently created. We then applied the biopsychosocial model to further characterize our findings. Participants included 26 patients and 11 caregivers. Patients were mostly middle-aged (mean age 37.7) males (80.8%), residing in urban settings (57.7%), injured in road traffic accidents (65.4%), and who required surgery (69.2%). Most caregivers were female. Seven major themes arose: pain, decreased physical functioning, poor emotional health, lack of support, challenges with daily activities, financial strain, and obstacles to accessing healthcare. This study describes some of the difficulties transitioning back into the community after hospitalization for traumatic injury. Our work demonstrates the importance of mixed methods approaches in characterizing and addressing transitions of care challenges.