Nosocomial tuberculosis: the blind spot in global tuberculosis prevention and care

Tuberculosis (TB) re-emerged as the world’s leading cause of death due to an infectious disease in 2023, with an estimated 10.8 million people falling ill and approximately 1.25 million deaths globally, surpassing COVID-19 in annual mortality [1]. Hospitals in high-burden regions are often sites for TB transmission, including drug-resistant Mycobacterium tuberculosis, despite knowledge about strategies that are effective in preventing nosocomial transmission [2,3]. Nosocomial transmission of M. tuberculosis is likely underrecognized because of diagnostic and surveillance gaps and competing demands on limited resources. Outbreaks of TB, including multidrug and extensively resistant disease have been repeatedly traced to healthcare facilities, endangering both patients and healthcare workers [4,5].

Whole genome sequencing (WGS) can unmask TB transmission chains, offering far better resolution than older or what had been conventional genotyping methods [6,7]. Yet, more than a decade after WGS became widely available, its use in high TB burden settings remains limited. WGS studies have been used in investigating M. tuberculosis transmission on multiple continents. To date, only a few prospective studies—most recently in Moldova [8] and now Kyrgyzstan [9]—have combined WGS with patient movement and room-sharing data to investigate nosocomial spread. Kyrgyzstan, with one of the world’s highest rates of drug-resistant TB, including multidrug and extensively drug-resistant TB, provided the setting for this current investigation [1].

In a longitudinal study conducted over 20 months in two TB hospitals in Kyrgyzstan, investigators followed 563 patients with culture-confirmed pulmonary TB and applied WGS to their baseline and follow-up M. tuberculosis isolates [9]. Two definite transmission events were documented corresponding to an incidence of 13.7 per 1000 patient-years, with wide CIs expected given the small number of events. Both index cases had drug-resistant TB, and the investigators documented M. tuberculosis transmission to patients sharing the same room who were hospitalized for treatment of drug-susceptible TB. An additional five potential transmissions were identified among patients who overlapped in hospital time but did not share rooms, which raised the combined estimate to 47.9 per 1000 patient-years, a markedly higher rate than previously reported in WGS-based hospital studies [7,8], although based on small numbers. As the authors acknowledge, their study almost certainly underestimates the degree of nosocomial transmission as the median follow-up was only about 90 days.

WGS analysis also showed that nearly one-third of baseline isolates clustered within ≤5 single nucleotide polymorphisms (SNPs), indicating a substantial level of recent community transmission that should inform public policy. Methodological strengths of the study included near-complete isolate collection, integration of patient room-sharing and temporal overlap data, and serial sampling over time, all of which allowed WGS to better characterize M. tuberculosis transmission dynamics. For a country with a high burden of drug-resistant TB, this level of surveillance is uncommon and provides a clearer view of nosocomial risks in TB hospitals.

The study has a few important limitations. Most notably, patients were followed for a median of only approximately 90 days and apparently only while hospitalized. This period is too short to capture all cases of the reinfection events, and there was no postdischarge surveillance, which means nosocomial transmission was almost certainly underestimated. The incidence rates are also based on a relatively small number of events, leaving wide CIs that make the estimates somewhat unstable. The analysis was further restricted to just two TB hospitals in Kyrgyzstan, limiting generalizability and making national extrapolations by the authors uncertain and likely unreliable. The supplemental material indicates that the United States Agency for International Development (USAID) Defeat TB project had invested in TB infection prevention and control (IPC) measures across Kyrgyz hospitals, but the authors do not describe what specific measures were implemented inside the two TB hospitals that were the focus of the study. Importantly, patients at these two TB hospitals routinely mixed in shared spaces—dining halls, smoking areas, and procedure areas making hospital-acquired transmissions of M. tuberculosis outside of shared rooms highly plausible and a likely explanation for the five events detected among patients who were not sharing rooms.

This study highlights the risks of continuing the Soviet-era model of TB care [10] that hospitalizes patients for prolonged periods, including throughout the intensive treatment phase that was the local standard of care at the time of the study. In the two Kyrgyz TB hospitals studied, patients with drug-susceptible TB were placed in the same wards—and sometimes even the same rooms—as patients with multidrug-resistant TB (MDR-TB), creating ideal conditions for superinfection and resistance amplification. The Krygyz study was conducted from 2017 to 2019, and there is no discussion of whether practices have evolved since then. The WHO recommends outpatient, community-based care and not prolonged inpatient hospitalization; this is not only safer from an infection prevention standpoint but more cost-effective and patient-centred [11]. MDR-TB transmission to persons with drug-susceptible disease is not new and has been documented elsewhere [12]. The nosocomial transmission highlighted in this study is not just a problem confined to Kyrgyzstan’s TB hospitals. In high-burden countries, a substantial proportion of patients hospitalized in general hospitals may have TB. For example, findings from KwaZulu-Natal, South Africa, further underscore this risk. A province-wide hospital survey revealed that nearly one-fifth of inpatients had active TB disease, including high proportions with MDR- and extensively drug-resistant (XDR)-TB [4]. The Kyrgyz study findings, combined with reports of outbreaks in general hospitals [12–14], demonstrate that hospitals continue to be efficient settings for the amplification of TB transmission, including the spread of drug-resistant strains, when effective infection prevention measures have not been implemented. Multiple studies have demonstrated that a hierarchy of TB IPC measures (administrative, engineering, respiratory protection) are effective in preventing the nosocomial transmission of TB, with administrative controls being the most important [2,3].

WGS remains underutilized in TB prevention and care activities especially in high TB burden low- and middle-income country settings, even though it consistently reveals transmission that conventional methods miss. Beyond the Kyrgyz study, WGS has revealed unsuspected nosocomial transmission in Southern Brazil, genomic analysis of XDR and pre-XDR M. tuberculosis strains identified hospital transmission clusters that were not evident through standard epidemiological investigation [15]. In France, prospective WGS surveillance identified otherwise unrecognized nosocomial TB transmission [5]. These examples demonstrate that nosocomial TB is not restricted to high-burden regions, and that WGS has become a powerful tool for uncovering hidden chains of transmission across diverse settings [16]. The Kyrgyzstan study adds to this evidence by combining longitudinal WGS analysis with room-sharing data and serial sampling, enabling unusually precise attribution of hospital-based transmission events. By sequencing M. tuberculosis from nearly all admitted patients, the investigators provided a more complete picture of transmission dynamics than traditional diagnostic or epidemiologic methods alone. Embedding WGS into routine surveillance activities should be a priority to advance the goals of the WHO End TB Strategy by guiding targeted interventions to disrupt transmission.

The Kyrgyzstan study highlights that nosocomial transmission of M. tuberculosis remains a serious and often under-addressed problem. The global community continues to have a blind spot for nosocomial TB and has not taken the necessary steps to prevent transmission of TB to patients and health care providers. Through the use of WGS, the investigators of the Krygyz study identified multiple transmission events occurring in hospital settings, including drug-resistant strains. There is no discussion of whether the findings of the study led to any changes in care or IPC practices. The study also reflects a broader global failure: we know what works—shorter hospital stays, administrative controls that include separation of patients with and without TB and in TB hospitals avoiding cohorting or comingling persons with drug-susceptible and drug-resistant TB; rapid diagnostics and initiation of appropriate anti-TB therapy; and improved ventilation—but these measures remain underutilized. This is in part because of the lack of political will necessary for sustained investment and prioritization of IPC in high-burden and often resource-limited settings [17]. The lesson from Kyrgyzstan emphasizes that WGS is useful in revealing nosocomial transmission and the need for actionable public health and IPC interventions to reduce the risk to patients and health workers, especially those in high-burden settings. As a global community, we cannot keep describing the problem while failing to intervene. Genomic evidence must be translated into practice, and IPC measures must move from the margins to the centre of TB prevention and care activities.