Implementation of a clinical decision support algorithm (CDSA) to De- implement unnecessary antibiotic prescriptions among HIV-infected adults with upper respiratory tract infections in primary healthcare settings in the cities of Maputo and Matola, Mozambique

Abstract

Antibiotics are commonly overused to treat upper respiratory tract infections (URTIs) in HIV-infected adults, even though viruses cause most URTIs. Therefore, antibiotic overuse for URTIs is considered an unwarranted and unnecessary practice, which needs to be de-implemented. Thus, the goal of this thesis is to evaluate the effectiveness and implementation of a clinical decision support algorithm (CDSA) to de-implement unnecessary antibiotic prescriptions among HIV-infected adults with URTI symptoms in primary healthcare facilities. The thesis comprises a total of four studies, culminating in five manuscripts. Study I (Paper I) aimed to explore and describe antibiotic prescribing for HIV-infected patients in primary healthcare facilities, using a cross-sectional design. Antibiotics were prescribed in 65.9% of prescriptions, either for treatment (69.8%) or prophylaxis (30.2%) of infections, and most were indicated for respiratory tract infections (30.5%). Study II (Paper III) aimed to explore the context of readiness for implementing the intervention in selected healthcare facilities, using a cross-sectional design with a mixed-methods approach. Thirty-nine healthcare providers (HCPs) among clinicians, laboratory technicians, and pharmacists were interviewed. Over 50% of clinicians did not possess or use any clinical guideline/algorithm, and 92.6% reported using clinical diagnosis alone to determine the choice in antibiotic use. All study facilities reported limited laboratory capacity to aid evidence to clinicians in antibiotic prescribing. HCPs described enthusiasm and willingness to utilize a new CDSA intervention. The lack of existing decision-support tools and limitations in laboratory diagnostic support justified the introduction of our CDSA, and the HCPs’ enthusiasm and willingness supported their readiness. Study III (Papers II and IV) aimed to evaluate the effectiveness of the intervention on reducing unnecessary antibiotic prescriptions for URTIs among HIV-infected adults, using a two- arm cluster randomized controlled trial design. Three hundred seventy-nine patients were recruited, comprising 182 (48%) in the intervention group and 197 (52%) in the control group. Most appeared with common cold and flu- like symptoms. The intervention was associated with a significant reduction in antibiotic prescribing by 33.2% (p < 0.001) and a non-significant decrease in incidence of complications by 3.7% (p = 0.096). In both groups, most patients (78%) recovered completely within five days. Amoxicillin (47.8%), azithromycin (21.9%), and phenoxymethylpenicillin (14.1%) were the most prescribed antibiotics. Study IV (Paper V) aimed to evaluate implementation outcomes of the intervention using the RE-AIM framework, employing a hybrid type II effectiveness-implementation design. Among 387 HIV-infected iv 7 adults approached, 379 (97.9%) were successfully recruited, with 182 (48%) in the intervention and 197 (52%) in the control group. Among the recruited patients, the mean age was 44±12.3 years, and 286 (75.5%) were female. The intervention resulted in 33.2% fewer antibiotics prescribed compared to the control. All intervention sites (100%) and clinicians (100%) demonstrated a commitment to de-implementing antibiotics. The implementation protocol was delivered as planned, and participants (n=21) in focus group discussions (FGD) were satisfied with the intervention. The evidence presented in this thesis may support clinicians and decision makers in their efforts for rational antibiotic use in managing URTIs in primary healthcare facilities