We conducted an exploratory cross-sectional study from September to December 2023, across eight hospitals in Zambia, namely, the Arthur Davison Children’s Hospital (ADH), Ndola Teaching Hospital (NTH), Kitwe Teaching Hospital (KTH) in the Copperbelt Province, Chilonga Mission Hospital (CMH) in Muchinga Province, Chipata Central Hospital (CCH) in the Eastern Province, Kabwe Central Hospital (KCH) in the Central Province, Livingstone Central Hospital (LCH) in the Southern Province, and Mansa General Hospital (MGH) in the Luapula Province (Figure 1). These healthcare facilities were chosen as they are the referral hospitals in the selected six out of the ten provinces in Zambia with high hospital annual admissions ranging from 3,500 to 19,656 patients. The number of bed spaces for these hospitals ranged from a minimum of 230 to a maximum of 741 as shown in Table 1. The facilities were also considered because of their ability to provide minimum to advanced laboratory testing to support AMS programs compared to other hospitals in the country. The selected hospitals were the sentinel sites for AMR surveillance in the country though they did not have established AMS programs. Therefore, the baseline assessment was conducted in these AMR surveillance sites to investigate their capacity to implement AMS programs. The participants of this study included domiciled healthcare workers who were members of the respective hospitals’ AMS, DTC, or IPC teams inclusive of pharmacists (25%), physicians (25%), environmental health technologists (9%), laboratory scientists (22%), and nurses (19%) as shown in Table 2. Although self-treatment and sale of antibiotics without prescriptions in informal and retail pharmacies are major contributors to the development of AMR, this study only targeted hospitals and excluded retail or informal pharmacies.

To get a representative sample, we interviewed four members of the AMS team, DTC, or ICC per healthcare facility. This translated into a minimum sample of 32 respondents. All participants were selected using a purposive sampling method because they were the best source of information regarding AMR activities in the selected healthcare facilities. Purposive sampling is useful when enrolling participants or institutions with specific characteristics (80). Therefore, we used a purposive sampling method to identify and enrol specific healthcare workers who are members of the DTC, ICC, or AMS committees. Additionally, these participants were from multi-disciplinary professions involving Pharmacists, Nurses, Clinicians, Laboratory personnel, and Environmental technologists.

Data were collected using the validated Periodic National and Healthcare Facility Assessment Tool in the WHO policy guidance on integrated AMS activities in human health (81). Using this tool, healthcare facilities can assess their level of preparedness for AMS about their core elements, create a step-by-step implementation plan, and track their progress over time in implementing AMS programs and activities (81). The tool was developed by the WHO to guide member states on how to implement AMS activities in a programmatic and integrated approach. It complements the Global Action Plan on AMR and the WHO toolkit to address AMR in low-and middle-income countries (82, 83). The assessment tool assesses healthcare facility AMS using six core elements with each having components of assessments (indicators). The core elements included; General [presence of DTC, Infection Control Committee (ICC), and AMS teams]; DTC functionality (9 indicators); Leadership commitment (6 indicators); accountability and responsibility (10 indicators); AMS Actions (11 indicators); Education and Training (3 indicators); Monitoring and surveillance (6 indicators); and Reporting Feedback within the Healthcare Facility (four indicators) (81).

The responses were scored as 0 = No, 1 = No but a priority, 2 = Planned but not started, 3 = Partially implemented, 4 = Yes (Fully implemented). The overall score of 80–100% indicated that AMS was fully implemented and functioning well, but needed continuous support for sustainability, 50–79.9%, AMS was partially functioning and needed attention for strengthening while 0.0–49.9% indicated that AMS was poorly functioning or non-functioning, and needed prioritized attention (81, 84). The tool was reviewed by members of the National AMS Technical Working Group (TWG) under Antimicrobial Resistance Coordinating Committee (AMRCC) at the Zambia National Public Health Institute (ZNPHI). No modifications were made to the tool.

The data collectors were also members of AMS National TWG. These were trained on how to administer the interview, sticking to the questions in the questionnaire, the importance of neutrality, and reducing bias. Additionally, role-playing exercises were carried out to help interviewers refine their ability to ask consistent and unbiased questions. Further, the three interviewers were staff from diverse backgrounds including nurses, clinicians, pharmacists, and laboratory staff. To familiarize the data collectors with the data collection tool, we conducted pilot studies in two hospitals in the Lusaka District which were excluded from the assessment. The face-to-face interviews were conducted by three data collectors per healthcare facility. All the data were collected using tablets and computers with restricted access to authorized personnel. Data collection took place for 2 days per hospital and proceeded without challenges.

Data were entered in the WHO self-scoring assessment tool in Microsoft Excel. Descriptive analysis was used to analyse the data using the WHO-validated self-scoring tool. The self-scoring tool provided a summary score of the six core elements of the assessment. This was followed by categorizing all responses into six themes under each core element including (i) Presence of DTC, ICC, AMS teams in surveyed hospitals, and DTC functionality, (ii) Leadership commitment, accountability, and responsibility, (iii) Antimicrobial stewardship actions, (iv) Education and training on AMS, (v) Monitoring and Surveillance, and (vi) Reporting feedback within the healthcare facility. Each of the core elements of AMS had an overall score that was calculated out of 100%. Interpretation of the findings was performed based on the Periodic National and Healthcare Facility Assessment Tool in the WHO policy guidance on integrated AMS activities in human health (81).

Ethical approval was obtained from the Tropical Diseases Research Centre (TDRC) Ethics Committee with an approval number of TRC/C4/09/2023. Permission to access the respective hospital was obtained from the Hospital Management Teams. Furthermore, informed consent was acquired from the hospital leadership and key informants. To preserve anonymity and protect hospital staff who took part in the study from any unfavorable outcomes, no personal identity information was collected.

This study found an average score of 56% for all AMS core elements in the priority healthcare facilities surveyed. MGH, KGH, LTH, NTH, and CCH had AMS programs partially functioning with scores of 56–78% and hence needed attention for strengthening while ADH, CMH, and KTH were poorly functioning, with scores of 41–44% and therefore, needed prioritized attention. The lowest scores on AMS activities were observed with ADH (41%), CMH (42%), KTH (44%), LTH (56%), and NTH (56%). The highest score in AMS core elements was recorded at MCH (78%) followed by KCH (67%). Therefore, no facility had an overall score above 80% to indicate a functioning AMS program. Overall, the assessment revealed the low performance of healthcare facilities on core elements such as DTC functionality (49%), leadership and commitment (56%), AMS actions (50%), education and training (54%), and reporting feedback within the healthcare facilities (46%) (Figure 2).

This indicator assessed the existence of the DTCs, AMS, or ICC and their functionality. Some of the indicators used to establish the functionality of these committees were the availability of regular meeting minutes and an antibiotic policy and procedure document, the presence of developed action plans that have been approved by management, and the categorization of antibiotics by the AWaRe classification.

Hence, the average score for the presence of DTC, AMS, and ICC was 91%, of which most facilities scored above 80%, with only two facilities scoring 75%. Nevertheless, the functionality of these committees was suboptimal with an average score of 49%. CCH and NTH did not have a fully functional IPC while ADH had no IPC committee (Supplementary Table S1). Similarly, ADH and Chilonga did not have AMS committees but had the DTC/MTCs (Supplementary Table S1).

Of the assessed facilities, none of the hospital DTCs had approved terms of reference, members assigned officially nor conducted supply and medicine use problem studies. Only KGH and MGH took action based on the findings from the supply and medicine study (scored 4).

KGH, CCH, and MGH had partially developed (score of 3) and none had fully developed DTC or AMS action plans (Figure 3). Whereas, ADH had planned for it (score of 2) but had not developed it. Three facilities (LTH, NTH, and MGH) had their specific health facility medicine lists categorized by AWaRe (score of 4) while the other three (ADH, CMH, and KGH) did not. Antimicrobial use policies and procedures were absent in all eight facilities with only CMH, MGH and CCH having partially developed (score of 3). However, facilities (KTH, LTH, NTH, KGH, ADH, and MGH) fully implemented hospital DTCs reporting their activities to management. Additionally, KTH, LTH, NTH, CCH, and ADH conducted regular meetings with documented minutes while CMH, CCH, and MGH partially implemented this. Further, four facilities (KGH, CMH, CCH, and MGH) had updated facility-specific medicines and medical devices while LTH and NTH partially implemented these indicators (Figure 3).

Leadership commitment to AMS involved prioritizing AMS programs by the leadership of the surveyed hospitals. This included supporting the DTC/AMS teams by allocating necessary human, financial, and information technology resources to AMS activities, as outlined in an endorsed institutional action plan. Whereas, accountability and responsibility involved the hospital having a single leader responsible for program outcomes on AMS. The leader or champion must have dedicated staff time for AMS activities in their terms of reference (TORs) or job description.

In this study, 50% (4/8) of facilities, namely ADH, KTH, CMH, and NTH, scored below average (56%) on leadership commitment to AMS activities (Figure 2). Nevertheless, AMS was considered a priority (score of 4) by the majority (6/8, 75%) of the healthcare facility management including NTH, KTH, LTH, CCH, MGH, and KGH. Four facilities (NTH, KTH, ADH, and CMH) had no AMS activities in their institutional annual action plans with key performance indicators. Additionally, there was no dedicated financial support for the healthcare facility AMS action plan in any of the assessed hospitals (Supplementary Table S1).

The presence of multidisciplinary AMS committee leadership in the healthcare facility with clear TORs was absent in four facilities (LTH, NTH, CMH, and ADH), present in two (MGH, KCH), and partially implemented in two facilities (KTH and CCH). Two facilities (MGH, and LTH) had a dedicated AMS leader/champion though none had dedicated staff time for AMS activities in their TORs or job descriptions. Additionally, two facilities (ADH, and CMH) reported not disseminating the AMS activity report to facility management and other healthcare facility team members (Supplementary Table S1).

AMS actions included target issues or activities undertaken by the hospital AMS program.

These were assessed based on the availability of updated Standard Treatment Guidelines (STGs), regular auditing of antibiotic therapy, conducting regular ward rounds on AMS, AWaRe classification of antibiotics in the hospital formulary, access to an information management system for data collection by the AMS team, and a written AMS policy requiring prescribers to indicate the reason for antibiotic prescriptions.

Overall, 25% (2/8) of the hospitals (ADH and CCH) scored below average (50%), while 50% (4/8, KCH, MCH, CMH, and KTH) scored 50% due to a lack of comprehensive AMS actions (Figure 2). Four facilities (score of 4) (CMH, LTH, NTH, and CCH) had standard treatment guidelines though not reviewed and updated with new evidence periodically. Further, advice or feedback from the AMS team was only fully accessible in three facilities (KGH, NTH, and MGH) and partially (score of 3) in three facilities (KTH, LTH, CCH). Only KGH and MGH conducted regular audits of specified antibiotic therapy or clinical conditions. Additionally, the AMS teams did not conduct ward rounds or other AMS interventions in selected departments across four facilities (KTH, LTH, CMH, and ADH). Besides, most facilities (88%, 7/8) namely LTH, KTH, KGH, CMH, CCH, MGH, and ADH did not have formularies that list antibiotics according to the AWaRe classification that requires approval from a designated person or team, thereby increasing the risks of irrational prescribing. Of concern, 62.5% (5/8) of the facilities (KTH, CMH, ADH, CCH, and LTH) did not have standardized prescription charts, medical records, patient folders, and transfer notes to support treatment and AMS activities. All the facilities except CCH performed well (score of 4) in ensuring access to laboratory and imaging services that supported AMS interventions. Besides, only three facilities (NTH, KGH, and MGH) had information technology services or other inventory control tools. Finally, 63%, 5/8 of facilities (KTH, NTH, ADH, CCH, and LTH) did not have a written policy that required prescribers to document the indications and antibiotics prescribed in a prescription chart or medical records (Figure 4).

The assessment was specifically centred on enhancing the inclusion of AMS programs in staff induction training and continuous professional development on AMS and IPC covering programs such as optimizing antibiotic prescribing, and dispensing.

Our findings revealed that 37.5% (3/8) of the facilities (KTH, CMH, and LTH) scored below average (54%) concerning the education and training of staff on AMS (Figure 2), indicating that most HCWs were not trained on AMS. Only two (ADH, and MGH) of the eight facilities partially included AMS programs such as optimizing antibiotic prescribing, dispensing, and administration in the staff induction training. Further, only two facilities (KGH, and MGH) fully offered continuous in-service training or continuous professional development on AMS and IPC to staff.

This was evaluated hinging on the following; the regular prescription audits, point prevalence surveys assessing the appropriateness of antibiotic prescribing, regular monitoring of shortages or stock-outs of essential antimicrobials, and compliance to the predetermined AMS intervention.

Therefore, based on the above indicators, three (KTH, CMH, and ADH) of eight hospitals did not conduct regular prescription audits, or point prevalence surveys to assess the appropriateness of antibiotic prescribing, whereas four facilities (LTH, NTH, KGH, and CCH) did so but not regularly. Impressively, 87.5% (7/8) of the facilities (KTH, CMH, KGH, MGH, LTH, CCH, and ADH) regularly monitored the shortages or stock-outs of essential antimicrobials. Two facilities (KGH and MGH) also fully monitored compliance with at least one specific AMS intervention.

On reporting feedback, the assessment was based on the presence of reports on the quantities of antibiotics purchased, prescribed, and dispensed to prescribers. Additionally, the assessment considered whether systems were in place to link the monitoring and reporting of healthcare-associated infections (HAIs), antimicrobial use, AMR, patient outcomes, and quality of care. Further, the study assessed if the AMS teams communicated the findings from audits or reviews of the quality or appropriateness of antibiotic use to prescribers. The study found that more than half, 62.5% (5/8) of the facilities (ADH, CCH, KTH, CMH, and LTH) had challenges with reporting feedback within the healthcare facility and hence, performed poorly during the assessment. Most of the facilities scored below 50% with the highest and lowest being MGH (60%) and the lowest ADH (28%) (Figure 2). Two hospitals (NTH, and MGH) exhibited reporting on the quantities of antibiotics purchased, prescribed, and dispensed to prescribers. These two hospitals (NTH and KGH) also had evidence of antibiotic susceptibility rates and key findings being shared with prescribers (scored 4). Further, the assessment indicated that AMS teams from six facilities (LTH, KTH, CCH, CMH, ADH, and NTH) did not communicate their findings from audits or reviews of the quality/appropriateness of antibiotic use to prescribers along with specific action points. However, none (score of 1) of the hospitals assessed had systems linking the monitoring and reporting of healthcare-associated infections (HAIs), antimicrobial use, AMR, patient outcomes, and quality of care (see Figure 5).

Some of the challenges faced by healthcare facilities to implement AMS activities included a lack of leadership commitment, funding for AMS activities, dedicated human resources, inactive AMS multidisciplinary teams, lack of updated STGs, policies and tools, laboratory commodities, and ineffective communication between the AMS committees, prescribers and other health workers as shown in Table 3.