Robust infection prevention and control (IPC) implementation in health facilities reduces tuberculosis (TB) infection (TBI) risks for health care workers (HCW), clients, and visitors.¹ A study in China found that regular IPC trainings and ultraviolet germicidal irradiation (UVGI) equipment maintenance were associated with lower TBI rates at the facility level.² Despite the evidence supporting TB prevention measures, poor adherence to IPC standards has been documented.^3–5 As a result, TBI remains a substantial risk for HCW. A systematic review estimated the pooled prevalence of TBI to be 54% among HCW in low- and middle-income countries.¹ A separate review specific to China found an average TBI prevalence of 52%.⁶

Within a health facility, TB IPC implementation fidelity may be heterogeneous, and some settings may require tailored mitigation strategies. Departments associated with higher risk of TB disease include inpatient TB departments and laboratories.¹ Multiple studies in China have observed higher TBI prevalence among HCW in inpatient departments.^4,7 One study found that the odds of TBI were 2.9 times higher among HCW in inpatient TB wards and 1.8 times higher among those working in respiratory wards compared to general medicine facilities, although TBI conversion rates were not significantly different between departments at follow up.⁷

Previous studies have found associations between IPC practices and HCW knowledge and attitudes. A systematic review found that HCW knowledge of TB was significantly associated with implementing TB IPC practices.⁸ In another study, HCW reported factors affecting IPC guideline adherence related to attitudes, including workplace culture, trust in personal protective equipment, and a sense of duty to clients.⁹ Another review noted that low managerial support and negative HCW attitudes may impact TB IPC compliance.¹⁰

TB Building and Strengthening Infection Control Strategies (TB BASICS) is a framework for identifying and prioritizing opportunities to improve TB IPC practices and policies in health care settings. The project includes IPC training, mentorship, guided planning to address suboptimal IPC practices, and periodic TB IPC risk assessments. TB BASICS was implemented in 9 health care facilities in China from 2017 to 2019. In addition to the standard project activities, a knowledge, attitudes, and practices (KAP) survey was carried out among HCW at participating facilities. A previous analysis of project data found substantial improvements in TB IPC implementation during follow-up¹¹ but did not evaluate differences between departments, highest- and lowest-performing IPC indicators, or KAP survey results.

Nine TB-designated health facilities in 6 provinces in eastern, central, and western China participated in TB BASICS. Baseline IPC risk assessments were conducted at all facilities and repeated quarterly for 18 months using standardized tools.¹¹ Each participating department within a facility was evaluated using standardized, department-specific assessment tools. The inpatient department assessment tool included 33 indicators, laboratory 20, and outpatient 34. Some indicators were tailored to better fit the department type or omitted if irrelevant. For this analysis, each indicator received 2 points for full implementation, 1 point for partial implementation, and no points if not implemented. Risk assessment teams could also designate an indicator as not applicable to the department assessed; in these cases, that indicator would not count toward the department’s score. Scores were calculated as the sum of points awarded, divided by maximum possible. Average scores were evaluated by department and by individual indicator. In the department-wise analysis, scores were first averaged at the facility level. Differences between departments were assessed using the Kruskal-Wallis rank sum test with an alpha level of 0.05. The indicator-wise analysis was approached similarly, with indicators assessed in each department averaged first by department type and then by facility.

Prior to baseline IPC risk assessments, a KAP survey was administered to all HCWs assigned to the departments assessed. The same survey was repeated at the time of the final risk assessment. The knowledge portion included 10 multiple choice questions. Results were converted into facility-level knowledge scores by summing the percentage of correct responses and dividing by the total number of questions. In the attitudes portion, respondents were asked to rate 13 statements on a scale of 1 to 4, with 1 representing strong agreement and 4 representing strong disagreement. About half of the questions were phrased as positive views or attitudes that would be expected to support IPC guideline adherence. To assess self-reported IPC practices, respondents were asked 28 multiple choice questions. For some questions, respondents could select multiple responses. The KAP survey did not undergo validity testing, but questions were based on extensive literature review.

This was a China CDC and US CDC cooperation activity, and the protocol was reviewed and approved by both sides, deemed not research, and was conducted consistent with applicable Chinese and US law and policy. For US CDC, nonresearch determination was defined by the US Department of Health and Human Services, Protection of Human Subjects 45 CFR §46.102(l).

Among the 9 facilities participating in TB BASICS, 57 departments received risk assessments during follow-up. Of these, 40% were inpatient, 18% were laboratory, and 42% were outpatient departments.¹¹

Risk assessments highlighted controls that demonstrated substantial improvement during follow-up, while others, such as respiratory isolation, remained persistently low scoring. Limited changes in knowledge scores, and mixed responses on self-reported IPC practices may suggest the need for more intensive staff training. Whiles scores in all departments increased from the baseline to final assessments, differences between departments remained. Greater attention may be needed to client-facing departments for a facility to achieve high IPC standards in all areas. The greatest improvements in laboratories were documented for the worst-performing areas at baseline, which were related to ventilation and UVGI. Proper UVGI moved from a score of 0 at baseline to a perfect score at the final evaluation, although this was the indicator most commonly marked not applicable in all department types. Natural ventilation may not have been appropriate in laboratory settings. Improvements in these areas are particularly notable, given that UVGI and natural and mechanical ventilation have previously been reported to reduce the risk of nosocomial TB transmission.¹²

Although some improvement was documented, sustained low facility-based risk assessment indicator scores related to client separation suggests perceived barriers to adopting these measures. Low implementation of fast-tracking and client separation have previously been documented in health care facilities in China.⁵ Limited space for client separation has been noted in other analyses as an influence on TB IPC implementation.¹³

Ventilation/negative pressure and management of persons with cough also emerged as themes among low-scoring indicators. While signage related to natural ventilation became more prevalent, the presence of openings on opposing walls remained low, likely due to the need for infrastructure changes. Keeping doors and windows open may have been impractical depending on weather conditions.

While some indicators assessing management of persons with cough did not achieve optimal implementation, 6 out of 11 indicators with perfect endpoint scores related to respiratory protection. This may suggest that implementation of respiratory protection controls related to supply availability and staff behavior were more feasible or preferable over controls aiming to affect client behavior.

While sputum-smear microscopy results could be obtained within 24 hours prior to TB BASICS implementation, timely availability of molecular diagnostic results was one of the least-improved indicators. However, this trend seems to have been driven primarily by outpatient departments, as this indicator was among the highest-scoring for inpatient departments throughout follow-up.

Full implementation of some indicators may have been affected by factors not assessed in this analysis. For example, a previous study in China found that having a designated sputum collection area was associated with having more staff and more clients with confirmed TB.⁵ An IPC measure such as outdoor sputum collection may also have been affected by the local climate.

Suboptimal implementation of some measures despite most respondents reporting recent TB and hospital-associated infections trainings suggests that annual periodicity may be insufficient. Relatively small changes in knowledge scores may indicate that additional refresher trainings were needed, or that staff attrition was high. Previous studies suggest that the relationship between TB IPC knowledge and routine practices is complex. Among HCW in Nepal, respondents working in TB wards with isolation rooms achieved higher TB IPC knowledge scores,³ demonstrating that levels of IPC knowledge may vary within department types. Studies in Brazil and Thailand found high TB IPC-related awareness but relatively low use of respiratory protection.^14,15

Respondents had generally positive views about their own capabilities and facilities’ willingness and ability to protect them. Overall, respondents tended to express that they had the appropriate IPC knowledge but lacked agency in improving IPC practices. A systematic review of HCW TB IPC perspectives found that respondents reported superior infection prevention practices than what was observed,¹⁶ which may help explain why self-assessments were high at the second KAP survey while gaps in TB IPC implementation remained. Almost all respondents reported that they would tell no one at work if they had TB symptoms, which may indicate fears of stigmatization.

Several limitations may have affected the results of this analysis. While the intention was to survey the same cohort at the repeat iteration of the KAP survey, staff attrition may have occurred in the interim. Question wording may have influenced responses. For example, the lowest-scoring knowledge questions related to TB transmission. Persisting low scores for these questions may have been due to use of terms such as “airborne transmission.” The potential for confusion around this term has since been acknowledged by a WHO Technical Consultation Group, which proposed use of “transmission through the air” to cover both airborne transmission/inhalation and direct deposition.¹⁷ While surveys were anonymous, respondents may have felt pressure to report what they considered were correct attitudes or practices. Differences in IPC scores by department may be a reflection of the applicable indicators in each setting. Laboratories are by nature more controlled environments than client-facing departments, which may help explain why laboratory scores were consistently higher than inpatient and outpatient departments.

This analysis, which follows an earlier manuscript describing the TB BASICS framework in China in greater detail,¹¹ suggests that IPC practices and policies may not be implemented uniformly within health facilities. Future IPC quality improvement initiatives may consider investing greater time or resources in outpatient departments. Ongoing efforts to continue improving TB IPC in health care facilities are needed to ensure adequate protection from nosocomial transmission for HCW and their clients.