http://repository.kemri.go.ke:8080/xmlui/handle/123456789/3 2026-08-26T03:33:50Z http://repository.kemri.go.ke:8080/xmlui/handle/123456789/1581 ACCEPTABILITY AND COST-EFFECTIVENESS OF HIV SELF-TESTING IN THE CONTEXT OF ASSISTED PARTNER NOTIFICATION SERVICES IN KISUMU AND HOMA BAY, KENYA. MUDHUNE, VICTOR OTIENO Assisted partner services (aPNS) is a recommended public health approach to promote HIV testing for sexual partners of individuals diagnosed with HIV. Offering HIV selftesting (HIVST) within aPNS has the potential to increase HIV testing uptake and yield, yet it is not well studied. This thesis evaluated acceptability and cost-effectiveness of HIVST within the aPNS-HIVST study conducted in western Kenya. The sociodemographic and behavioral characteristics of aPNS identified sexual partners offered the choice of HIVST or provider-delivered testing were evaluated, with a sub-set of partners purposively selected for in-depth interviews (IDIs) (n=24). In addition, focus group discussions (FGDs) with HIV testing service (HTS) providers were conducted (n=2), including IDIs with pharmacy staff involved in dispensing HIVST to aPNS clients (n=6). Descriptive and log-binomial regression analyses were performed controlling for health facility clusters, while interviews were thematically analyzed applying the theoretical framework of acceptability (TFA). Further, we conducted micro-costing, time-and-motion, and provider surveys to determine incremental HIVST distribution cost. Using a decision tree model, we estimated the incremental cost per new diagnosis for HIVST incorporated into aPNS, compared to aPNS with provider-delivered testing only. Uptake of HIVST was determined to be 82.2%, with no association between partner demographics and HIVST uptake. HIVST use was less likely than provider-delivered testing among those identified as a casual (adjusted relative risk (aRR) = 0.93; 95% Confidence Interval (CI) 0.88-0.98) or transactional (aRR = 0.90; 95% CI 0.87-0.94) partner compared to those in a defined relationship. HIVST use was slightly lower among those offered the option of an additional kit when compared to those only offered one kit (aRR = 0.93; 95% CI 0.88-0.98). In the partner IDIs, HIVST was a viable option for individuals who do not find provider-delivered testing suitable or convenient. For them, ‘intervention coherence’, ‘self-efficacy’, and ‘ethicality’ presented as most significant TFA constructs. aPNS providers played a critical role in creating HIVST awareness and driving acceptability, and cited benefits such as increased efficiency in conducting HIV testing. They reported challenges including maintaining confidentiality, delivering procedures remotely, HIV-related stigma at dispensing pharmacies and low HIVST awareness. Successful integration was facilitated by training and motivation of HTS providers to focus on benefits of HIVST to clients. Unexpected effects of HIVST introduction included increased HIVST community awareness and testing at the dispensing pharmacy, with resulting concerns about dispensing staff’s ability to deliver counselling effectively. The cost per HIVST distributed within aPNS was $8.97, largestcomponent being testing supplies (38%) and personnel (30%). Under conditions of facility-based testing uptake of <91%, or HIVST utilization rates of <27%, HIVST integration into aPNS is potentially cost effective. At willing-to-pay threshold of $1,000, the net monetary benefit was sensitive to effectiveness of HIVST in increasing testing rates, phone call rates, HIVST sensitivity, HIV prevalence, cost of HIVST, space allocation at facilities, and personnel time during facility-based testing. In a best-case scenario, HIVST option was cheaper by $3,037 and diagnosed 11 more cases. Increasing HIVST awareness and providing tailored solutions will empower aPNS clients optimize their HIV testing decisions. Providers should consider context of the partner’s sexual encounter and extend counselling support when recommending HIVST within aPNS. The integration should carefully consider four components of the five ‘Cs’ of HTS: confidentiality, opportunity for counselling, correct results, and linkage to care. In addition, pharmacy staff dispensing HIVST must receive HTS training, and HTS providers must be trained on remote telephone delivery of counselling and aPNS procedures. Implementers and policy makers should ensure HIVST efficiency by focusing on facilities with low uptake for provider-delivered facility-based testing, while deliberately promoting utilization among those likely to benefit from remote testing. Additional measures should focus on minimizing costs relating to personnel and testing supplies. 2025-01-01T00:00:00Z http://repository.kemri.go.ke:8080/xmlui/handle/123456789/1580 Utilization of a Targeted Mobile Phone Intervention in Antenatal Care and its Effects on Postnatal Outcomes Among Pregnant Women in a Pastoralist Community in Narok County in Kenya Muvengei, Daniel Mwendwa Introduction Worldwide, complications that occur in pregnancy, at childbirth, and in the immediate postnatal period (puerperium) are the leading causes of mortality and morbidity among Women of Reproductive Age (WRA). Most women die because of complications occurring during and following pregnancy. The majority of these mortalities and morbidity are due to preventable causes and almost all of them (99%) occur in low and middle-income countries (LMICs). Antenatal care (ANC) is one of the key high-impact strategies to improve maternal and child health globally. WHO recommends eight visits. In LMICs, only about half of pregnant women receive the WHO-recommended minimum ANC visits. Good quality ANC care can reduce maternal morbidity and mortality and perinatal morbidity. Using technology in maternal health (mHealth) improves outcomes. Methods This study was conducted in four busy public hospitals in Narok County. A Randomized Controlled Trial (RCT) was conducted to determine the effect of a targeted mobile phone intervention on antenatal and postnatal clinic attendance, level of skilled attendant delivery, and also the resultant postnatal outcomes. The study population comprised WRA who were expectant recruited early in pregnancy. Two hundred and eighty mothers (280) were recruited. Recruitment began in June 2018 and the study closed in March 2021. The intervention was bi-component, consisting of a standardized Short-message Service (SMS) sent fortnightly and a phone call reminder made one week (7 days) before the date the study mother had been booked to attend the ANC clinic. All study mothers were followed up from recruitment to 42 days post-delivery. Data were analyzed with Stata v14 using descriptive and inferential statistics. Ethical approval was obtained from SERU at KEMRI. Results Two hundred and sixty-two mothers completed the study giving a 93.6% completion rate. The mean age at enrolment of the study participants was 23.87 years (SD 5.22, 95% CI 23.23-24.50) with the youngest study participant being 14 years old and the oldest mother being 44 years old. Eighty-six percent (86.26% (n=226)) of the study participants were married. ANC Attendance The mean number of antenatal visits done by each study participant was 3.48 visits (SD 1.06, 95% CI 3.35-3.61). The intervention was associated with improved antenatal care clinic attendance amongst the study population with the mean number of antenatal visits being 4.099 visits for the 131 study participants in the intervention group while it was 2.843 visits for the 128 study participants in the non-intervention study arm giving a difference in means of 1.256 visits (95% CI 1.044-1.467, p-value < 0.0001) Skilled Care deliveries The intervention was associated with improvement in skilled care deliveries. In the intervention group with 130 study participants, 90% (n=117) of them were assisted by a healthcare worker to deliver. Five percent (5.38%, n=7) of the study mothers were assisted by a relative to deliver while 4.62% (n=6) were assisted by a traditional birth attendant. In the non-intervention study arm with 128 study participants, 69.53% (n=89) were assisted by a healthcare worker to deliver while 15.63% (n=20) were assisted by a relative to deliver with 14.84% (n=19) being assisted by a traditional birth attendant. The null hypothesis of there being no difference in the likelihood of a study mother being assisted to deliver by a healthcare worker by the study group, was rejected (X2 16.810, p-value < 0.0001) indicating that the study mothers in the intervention study arm were more likely to be assisted by a healthcare worker to deliver than those in the nonintervention study arm. The intervention was also associated with improved health facility-based deliveries. Of the 130 study mothers in the intervention study arm, 90% (n=117) delivered in a health facility while 10% (n=13) delivered at home. Of the 128 study mothers in the non-intervention study arm, 69.53% (n=89) delivered in a health facility while 30.47% (n=39) delivered at home. The difference of proportions for the health facility-based deliveries between the intervention and non-intervention study arms was 20.47% (95% CI 10.97-29.96, p-value < 0.0001), which was statistically significant. Postnatal Clinic (PNC) Attendance The intervention was also associated with improved postnatal care clinic attendance. Of the 60 study participants who were in the intervention study arm, 41.67% (n=25) attended the PNC while 58.33% (n=35) did not attend. Of the 76 study participants who were in the nonintervention study arm, 21.05% (n=16) attended the PNC while 78.95% (n=60) did not attend. The null hypothesis that there was no difference in the likelihood of a study mother attending postnatal clinic by study group was rejected (X2 6.766, p-value 0.009). Maternal Postnatal Outcomes The intervention was associated with fewer maternal complications during and after delivery. The likelihood of a study participant having any complication at birth was 33.33% (n=86) while that of a study mother having no complication was 66.67% (n=172). The difference in proportion between the study participants who had a complication at birth was 17.23% (95% CI 6.51-27.94, p-value = 0.002) between the intervention (19.70%) and the non-intervention (36.92%) study arm which was statistically significant. Neonatal Outcomes The intervention was also associated with fewer neonatal complications. The median APGAR score at 5 seconds for the 121 study mothers in the intervention study arm was 10 while it was nine for the 99 study mothers in the non-intervention study arm. Mann Whitney U test of significance was used to test the null hypothesis that there was no difference between the median APGAR score at 5 seconds of the two study groups. The difference between the medians was found to be statistically significant at a 95% confidence level (Z statistic of 4.698, p-value < 0.0001). A targeted mobile phone intervention was associated with fewer neonatal mortalities. The difference in proportion between the study participants who had neonatal mortality at birth was 9.32% (95% CI 1.91-16.74, p-value = 0.015) between the intervention (6.06%) and the nonintervention (15.38%) study arm which was statistically significant. In conclusion, a targeted mobile phone intervention used in antenatal care was associated with improved ANC and PNC attendance and better maternal and neonatal postnatal outcomes. 2025-01-01T00:00:00Z http://repository.kemri.go.ke:8080/xmlui/handle/123456789/1579 Effectiveness of an Enhanced Community Health Volunteers Model in Improving Screening and Linkages for Non-Communicable Diseases in Nyeri County, Kenya Langakuo, Kenneth Mugambi Non-communicable diseases (NCDs) account for 63% of deaths worldwide. Of these, fifty per cent of hospital admissions and over eighty per cent occur in LMICs. The Kenyan government prioritized communicable disease management more than NCDs. Consequently, this affected the ill-equipped health systems, resulting in an upsurge in NCD morbidity and mortality. The study assessed how a modified CHV model worked in Nyeri County for NCD screening and linkages. This study utilized a mixed-method quasi-experimental study with a non-equivalent pre- and post-test method. Ten community units were sampled using a multi-stage cluster sampling technique. Each unit comprised 30 CHVs systematically chosen to participate in the study, with 150 CHVs in the interventional arm and 150 CHVs in the control group. Ten CHVs were purposively selected to participate in Key Informant Interviews, and six FGDs comprised of 10 discussants were carried out among selected community units. Data was collected and stored in the Kobo App, and later, quantitative data was analyzed using SPSS version 26.0. Interviews and group discussions were audio-recorded, and verbatim transcription and content thematic analysis were done via N-Vivo software version 12. The study showed that n=184 (61.3%) were females and n=218(72.7%) were aged 40 years and beyond, with n=215(71.7 %) having achieved a secondary level of education. Further, the study demonstrated that screening and linkage were higher in the intervention group (M =33.12, SD =2.50) (p =0.015). There existed a significant disparity in capacity from pre-intervention 48.75 % (SD±5.7) to 68.28 % (SD±7.6) and p <0.001. Physical activity contributed to NCD-related mortality (p =0.609) with notable behavioral modifiable risk factors (p=0.102). CHVs in the intervention group had a higher understanding of their role in the health workforce (M =3.49, SD =0.64) (P =0.010). Notably, CHVs in the intervention group understood that they work voluntarily (M =3.41, SD =0.59) (p =0.001). As part of their routine work, CHVs in the intervention group (M =3.27, SD =0.50) monitored their client's progress (p <0.001). Therefore, CHV training and orientation are essential to enhancing this capacity (P=0.001). The interventional group scored higher (p<0.001) concerning community knowledge and perceptions of roles and equally scored higher (P=0.002) in integrating with local structures. In addition, CHV perception of roles and influence was high in the intervention group (M =4.27, SD =0.49) (p<0.001). Average knowledge scores were higher in the intervention group (70%) p <0.001, and perception and roles about CHVs have influenced their performance (P=<0.001). All discussants highlighted traditional beliefs, culture, religion, myths, workload and inadequate training, together with misconceptions about their role and NCDs, as significant barriers and challenges to practical CHV work. However, gender and poor communication were significantly highlighted as barriers. The findings highlight that CHVs play a crucial role but need optimization through leveraging skills training and embracing community-based intervention and strategies. Therefore, continuous capacity building, innovative mobile app development and utilization, reinforced policy formulation together with communication, social mobilization, equitable resource allocation and community engagement programs will improve NCD screening and linkages, which later translates to timely diagnosis, treatment, referrals, promote care and add knowledge through publications in peer-reviewed journals and articles. 2025-01-01T00:00:00Z http://repository.kemri.go.ke:8080/xmlui/handle/123456789/1578 Epidemiological and Genomic Patterns of Methicillin-Resistant Staphylococcus Aureus Isolates Obtained from Patients Seeking Treatment in Nakuru County Referral and Teaching Hospital, Kenya Chelang’at, Soi Joan Staphylococcus aureus is a pathogenic bacterium that can evolve rapidly to methicillin-resistant Staphylococcus aureus (MRSA). Methicillin-Resistant Staphylococcus aureus is a global health concern that is associated with significant morbidity and mortality. Several factors have been associated with MRSA which include antibiotic use, prolonged hospitalization and surgical intervention among others. The objective of this study was to determine epidemiological and genomic patterns observed from MRSA isolates obtained from patients seeking treatment in Nakuru County Referral and Teaching Hospital. This was a cross-sectional study with a laboratory experimental component conducted over six months in Nakuru County Referral and Teaching Hospital, Microbiology Laboratory. This study used routinely collected laboratory samples such as blood, urine, pus and CSF of patients seeking treatment in the hospital. The samples were selected using systematic random sampling until the 354-sample size was achieved. In addition, the demographic and clinical factors were collected from laboratory records while patient level factors were collected from the hospital records. Microbiology samples for culture and sensitivity received from inpatient and outpatient were included in the study. Methicillin Resistant S. aureus isolates were identified phenotypically and the MRSA strains were confirmed using Whole Genome Sequencing. This data was then analyzed using IBM SPSS software Version 23. The association between independent variables and the S. aureus and MRSA were tested using Chisquare test for categorical variables and Fishers exact test as an alternative in case of small values. Mann Whitney U test was used to compare median ages. Statistical significance was interpreted at 5% level (p< 0.05). The study identified a prevalence of 7.1% for Staphylococcus aureus and 2.3% for MRSA among the sampled patients, with the highest prevalence of 28.6% observed in the burns unit. MRSA isolates exhibited significant resistance to penicillin G, erythromycin, and clindamycin, with resistance rates of 64% for these antibiotics. Despite this high resistance, MRSA isolates were highly susceptible to Cefoxitin (68%), Doripenem (56%) and Sulfamethoxazole (68%). Molecular typing revealed the dominance of clonal complexes CC8 and CC15, with ST152 being the most prevalent strain. A significant proportion (52%) of the isolates harbored virulence genes such as lukF-PV and lukS-PV, which are associated with severe infections. Key patientlevel risk factors identified included prolonged hospital stays and antibiotic use, which significantly increased the risk of MRSA infections. The study highlights critical epidemiological and genomic characteristics of MRSA at Nakuru County Teaching and Referral Hospital. These findings emphasize the need for continuous MRSA surveillance, stringent infection control practices, and routine antibiotic susceptibility testing to guide effective treatment. Understanding the genetic characteristics and virulence factors of MRSA isolates can aid in predicting clinical behavior and developing targeted interventions for managing and preventing MRSA infections. The recommendations from this study emphasize the critical need for robust infection control measures, regular surveillance of MRSA prevalence and resistance patterns, and prudent use of antibiotics to combat the spread of resistant strains. Enhancing laboratory capabilities, educating healthcare workers, and fostering interdisciplinary collaboration are vital steps in managing MRSA. Further research and public health initiatives are essential to raise awareness and develop effective preventive measures. Findings from this study was disseminated to the hospital's Medical Superintendent and healthcare workers, aiding in the proper management and prevention of MRSA infections. 2025-01-01T00:00:00Z