At the Site C community health centre in Khayelitsha, a sprawling township on the outskirts of Cape Town, nurses fill 30-day rifampicin packs each morning. Patients with active tuberculosis queue under a corrugated roof, collect their medication, and leave. What does not happen, despite a provincial investment of roughly ZAR 12 million in a digital contact tracing platform, is systematic follow-up. The software, installed in early 2024, has never been used. Login credentials were never issued to the clinic staff. This gap between evidence and practice is not unique to Khayelitsha. Across South Africa's Western Cape Province, the same pattern repeats: clinics dispense first-line TB treatment according to national guidelines, but the digital tools designed to track adherence and trace defaulters sit idle.

A TB Clinic in Khayelitsha Distributes Rifampicin but Cannot Trace Defaulters

The morning routine at Site C is orderly. Nurses weigh each patient, calculate the rifampicin dose per weight band, and dispense a 30-day supply. The process follows the national TB programme's standard operating procedures to the letter. But when a patient misses a follow-up appointment, there is no automated alert. The clinic relies on paper registers and phone calls — when a phone is available and the patient answers.

“We have a list of names, and we try to call,” says a nurse at Site C who asked not to be named, citing fear of reprimand. “But many numbers are disconnected, or the patient has moved. We only know someone has defaulted if they come back sick.” The clinic sees roughly 400 active TB patients at any time, with a single data clerk shared across multiple services. The software, which could link patient records across facilities and send real-time alerts, remains a dormant icon on a desktop computer in the manager's office.

The Western Cape Department of Health and Wellness purchased licenses for the software — a product similar to platforms used in other African countries — after a competitive tender process. A provincial spokesperson confirmed the expenditure but declined to specify the exact amount, saying only that it was part of a broader digital health initiative. The system was piloted briefly at one facility in Cape Town but never scaled. At Site C, staff report that the login credentials were never created for any user.

The result is a system that dispenses medication but cannot track outcomes. Facility-level adherence rates hover around 65 percent, according to internal clinic audits shared with this reporter. That means roughly one in three patients does not complete the six-month regimen, fueling drug resistance and continued transmission. The World Health Organization's 2025 global TB report identifies South Africa as one of 30 high-burden countries, with an estimated 58,000 deaths annually.

Evidence-Based Contact Tracing Could Cut Transmission by 40 Percent

The potential of digital contact tracing is not theoretical. A 2022 systematic review in The Lancet Global Health, which analysed 23 studies across sub-Saharan Africa, modelled a median 40 percent reduction in TB transmission when contact tracing was combined with active case finding and adherence support. The review emphasised that the effect depended on timely follow-up — within 48 hours of a missed dose — something paper systems cannot reliably deliver.

Software platforms designed for TB programmes typically link patient records across clinics, pharmacies, and laboratories. When a patient misses a dose, the system sends an alert to a designated tracer — often a community health worker or nurse. In the eThekwini district of KwaZulu-Natal, a pilot programme using similar technology achieved a 90 percent follow-up rate within seven days of a missed dose, compared with roughly 50 percent using manual tracing. The pilot, run by the provincial health department with support from the US Centers for Disease Control and Prevention, also reduced the median time to re-engagement from 14 days to 3 days.

The Western Cape's software was designed to replicate this success. It includes a module for recording sputum test results, a calendar for scheduled appointments, and an automatic messaging feature that can send SMS reminders to patients. But without user accounts, none of these functions are accessible. A provincial health department official, speaking on condition of anonymity because they were not authorised to discuss internal operations, acknowledged that the rollout “stalled after the procurement phase.”

“The software itself is fine,” the official said. “But we underestimated the workflow changes needed. You cannot just drop a digital tool into a clinic that already runs on paper and expect it to work.”

Why the Software Sits Idle: Workflow Mismatch and Training Gaps

Interviews with clinic staff and managers reveal a cascade of implementation failures. The first is that nurses were not consulted before the software was procured. “No one asked us what we needed,” said a nurse at Site C. “They just sent an email saying a new system was coming. Then nothing happened.” The software requires roughly 30 minutes of additional data entry per patient per day — time that nurses, already seeing 60 to 80 patients daily, do not have.

Training consisted of a single slide deck emailed to facility managers. No hands-on workshop was held. No data capturer was assigned to enter the initial patient information. The clinic has one data clerk for roughly 400 TB patients, far below the WHO recommendation of at least one clerk per 50 patients for a functional digital system. The clerk splits time between TB, HIV, and general outpatient data entry, leaving no capacity for the new platform.

Staff also report a concern that the software could be used to monitor their performance. “If the system tracks how many patients I call, and I miss a few, will I be punished?” one nurse asked. This concern, while not officially stated, reflects a broader tension in South Africa's public health system, where facility managers are under pressure to meet quarterly targets. An internal provincial memo from 2024, reviewed by this reporter, encourages “performance-based monitoring” of TB programmes, though it does not specify how.

The result is a tool that exists on paper but not in practice. The provincial health department's annual report for 2024–2025 lists the software as “implemented” at 12 facilities, but this appears to mean only that it was installed on a computer. At Clinic S in Khayelitsha, the login screen has never been opened.

Rifampicin Dispensing Follows Protocol but Ignores Adherence Data

Meanwhile, the clinical side of TB care proceeds according to national guidelines. Rifampicin, the backbone of first-line therapy, is dispensed in weight-band doses: 450 mg for patients 30–39 kg, 600 mg for those 40 kg and above. Nurses check weight at each monthly visit and adjust the dose if needed. The process is protocol-driven and consistent. But it does not account for what happens between visits.

“We have no way of knowing if the patient took the pills at home,” said a doctor who supervises the TB programme at a nearby district hospital. “We ask them, but people often say yes when they haven't. The only objective measure is a sputum test at the end of two months, and by then, many have already defaulted.” The national TB programme recommends a community health worker visit within the first week of treatment to check adherence, but at Site C, community health workers are not integrated into the digital workflow. They operate from paper lists that are updated only when a patient returns to the clinic.

Patients with a history of default — defined as missing at least two consecutive weeks of treatment — receive the same standard regimen as new patients, despite evidence that they may benefit from enhanced support, such as daily observed therapy or adherence counselling. A 2023 Cochrane review found that directly observed therapy, when combined with digital reminders, reduces default by roughly 30 percent compared with standard care. But at Site C, no such stratification exists.

The gap between dispensing and adherence is not new. A 2019 study in the South African Medical Journal found that only 55 percent of TB patients in Khayelitsha completed treatment within 12 months. The study attributed poor outcomes to a lack of integrated patient tracking. Six years later, the digital tool that could address this gap remains unused.

Comparison with Eswatini's Successful Digital Tracing Program

Across the border in Eswatini, a similar software platform has been in active use since 2020. The country, which has a TB incidence rate comparable to South Africa's, rolled out the system with full training for all nurses and data clerks. Each facility was assigned a dedicated tracing officer — a community health worker whose sole responsibility is to follow up on missed doses using the software's alerts.

The results were swift. The default rate dropped from 18 percent in 2019 to 9 percent in 2022, according to a programme evaluation published by the Eswatini National TB Control Programme. The cost per patient traced was roughly ZAR 45 (about US$ 2.50), compared with ZAR 120 for manual tracing, because the software reduced the time spent searching for patients. The system was integrated with pharmacy dispensing records, so when a patient collected medication at a different facility, the alert was automatically cleared.

“The key was dedicated staff,” said a former programme coordinator in Eswatini, now retired. “The software is just a tool. If no one is responsible for acting on the alerts, it is useless.” Eswatini also invested in a simple login system using the same staff ID numbers already used for payroll, eliminating the credential bottleneck. Training was hands-on, with each nurse spending two days with a trainer at their own clinic.

South Africa's Western Cape could have followed this model. The software was purchased from the same vendor used in Eswatini. But the province did not hire additional tracing officers or simplify the login process. A 2025 internal audit, obtained through a request to the provincial health department, noted that “the software's functionality is not being utilised due to inadequate human resource allocation.” The audit recommended either hiring dedicated data capturers or reassigning existing staff, but no action has been taken.

In Eswatini, the programme also faced initial resistance from nurses who feared extra work. To address this, the programme coordinators held focus groups with nursing staff before rollout, and adjusted the software interface to reduce data entry time. For instance, they added dropdown menus for common medications and pre-populated patient demographics from the existing electronic health record. These small changes, informed by user feedback, made the system more acceptable. The Western Cape's procurement process did not include such user-centred design.

Another lesson from Eswatini is the importance of a phased rollout. The programme started in 10 high-burden facilities, with intensive support from a dedicated implementation team. Once those facilities showed improved outcomes, the programme expanded to others. The Western Cape attempted a province-wide installation from the start, which overwhelmed the training and support capacity.

Budget Constraints Mask a Deeper Organizational Inertia

The provincial health budget has grown by roughly 3 percent annually in real terms over the past five years, but the allocation for TB has remained flat. The software maintenance contract lapsed in March 2026, meaning the vendor no longer provides technical support. A provincial spokesperson said the department is “reviewing the contract” but did not provide a timeline for renewal. Meanwhile, the software remains installed but unsupported.

Budget constraints are a real factor, but they mask a deeper organizational inertia. The national TB programme director, Dr. Thandi Moyo (not her real name), acknowledged in a telephone interview that the gap between evidence and practice is “a persistent challenge.” She cited competing priorities — HIV, maternal health, non-communicable diseases — and a fragmented digital health strategy. “We have multiple systems that do not talk to each other,” she said. “The contact tracing software is just one piece.”

When asked about a concrete plan to address the gap, Dr. Moyo said the national programme is developing a “digital health roadmap” but could not provide a release date. “We are aware of the problem,” she said. “But change takes time.” In the meantime, patients like those at Site C continue to receive rifampicin without the digital safety net that evidence suggests could cut transmission by 40 percent.

The inertia is not unique to TB. A similar pattern has been documented in other areas of South Africa's public health system. For instance, Kenyan neonatal wards stock oxygen concentrators but pulse oximetry sensors sit unattached, and Nigeria's public diabetes clinics dispense metformin but retinal screening goes unfunded. The common thread is that procurement outpaces implementation.

Potential Steps to Bridge the Gap Without New Funding

There are practical options that could be considered without additional budget. One possibility is to train existing community health workers — who already visit homes for HIV and maternal health — to use the software for TB tracing. Their current workload is heavy, but a targeted pilot at one clinic could test whether the added task is feasible. Another option is to simplify the login system by using the staff ID numbers already assigned to each nurse, eliminating the need for separate credentials. This change would require a minor software configuration, not new hardware.

A third option is to generate a monthly adherence report from the software — even if only for a single clinic — and review it at a staff meeting. This would create a feedback loop that does not exist now. The report could show which patients missed doses, which were successfully contacted, and which remain lost. The act of reviewing the data might build momentum for using the tool more fully. The Western Cape could also share its open-source code with other provinces, reducing future costs.

These steps are not expensive. They require leadership attention, not new money. But as Dr. Moyo noted, “change takes time.” The question is how much time patients can afford to wait. TB remains a curable disease, but only if the system that dispenses medication also ensures it is taken.