A descriptive evaluation study was conducted to assess the performance of MDR-TB surveillance system in Yemen using the Updated CDC guideline for evaluation of public health surveillance system [8]. The study was conducted at Sana’a city from November to December 2016. Data on MDR-TB were collected from the NTCP and the four Regional MDR-TB centers.

The MDR-TB is a sentinel system through four regional MDR-TB centers at Sana’a City, Aden, Taiz, and Al Hodaidah, with an aim to cover the whole country. The centers provide detection, diagnosis, treatment, and follow-up through a community-based outpatient treatment-delivery strategy that consists of a community supporter, community nurse, drug provider, coordinator, and lab technician. The community supporter is one of the patient’s family members who observes the patient daily, has contact with the community nurse regularly, and fills monthly reports for daily treatment intake. The community nurse follows the community supporters on a monthly basis, directly contacts patients during the follow-up visits, and is responsible for registration of DR-TB. The drug provider is a nurse responsible for supplying drugs to patients with MDR-TB and for the drug supply registry. All patients have monthly clinical assessments by a specialist doctor. The lab technician is responsible for performing the smears and cultures or others lab tests in addition to completing the lab registry and reports. The MDR-TB coordinator is responsible for sending the reports to the NTCP. All centers have equipped labs for drug susceptibility testing (DST) and culture. Due to a total breakdown of the electricity network that is needed for DST, the regional MDR-TB centers were provided with the GenXpert, except for the Taiz MDR-TB center, which still sends samples to Sana’a City for diagnosis.

At the MDR-TB centers, there are three hard-copy registers: the DR-TB case register, the lab register, and the MDR-TB drug supply register. The DR-TB cases register contains a set of variables including demographic variables, previous TB registration number, type of drug resistance, starting category 4 treatment date, and use of previous second-line drugs, DST result, smear and culture follow-up results, and HIV status.

The MDR-TB documents were reviewed to describe the system. The mangers of the NTCP were interviewed to assess usefulness, flexibility, and stability of the MDR-TB system. Semistructured questionnaires were used to collect data from staff of the four MDR-TB centers to evaluate simplicity and acceptability of the system. Specific items were used to evaluate each performance attribute. Respondents from the four centers were requested to rate the items measuring the system usefulness as well as the simplicity, flexibility, acceptability, and stability using the 5-point Likert scale (1=strongly disagree, 2=disagree, 3=not sure, 4=agree, 5=strongly agree). The mean percentage score was calculated for each attribute, where a higher score indicates a higher level of performance. The performance was interpreted using the following scoring system that was used in previous similar studies: <60%, poor; 60%-80%, average; and >80%, good [10-12]. To assess the data quality, we reviewed 10% of randomly selected MDR-TB cases’ files and registers. The selected files were reviewed for any missing data or discrepancies between data on cases’ files and register. Epi info version 7.2 (Centers of Disease Control and Prevention, Atlanta, Georgia) was used for data entry and analysis.

Figure 1 shows the data flow of the MDR-TB surveillance system, where the suspected case is sent to the nearest regional DR-TB center for lab confirmation and treatment. According to the tests results, the lab sends immediate reports for positive cases to the NTCP to be included in their drug supply for the reporting regional DR-TB. Each regional DR-TB center sends quarterly reports to the NTCP regarding MDR-TB cases during that quarter including lab report, drug supply, and treatment outcome. The NTCP sends quarterly and annual reports to the WHO.

The NTCP managers agreed that the data are useful. They reported that they used the data to estimate needs of the NTCP in terms of drugs and lab tests. Furthermore, all managers agreed that the MDR-TB data are used to estimate the TB magnitude, incidence, and mortality rates in order to monitor the trend of MDR-TB over time; identify areas at greater risk; update and develop the strategic direction for MDR-TB activity; and plan the resources for detection, prevention, and control activities. The usefulness percentage score was 87%, which indicates that the system is useful (Table 1).

All NTCP managers agreed that the system is able to adapt any change in case definition. Four managers (80%) reported that the system is flexible and can be integrated with other surveillance systems. For example, the system is integrated with the HIV program, where all MDR-TB cases are tested for HIV, and positive cases are reported to the HIV program. The flexibility score of the system was 61%, indicating average flexibility (Table 2).

All participants said that unscheduled outages have been rarely occurred before the war. Three (60%) managers said that the system would collapse if the donor fund stops. The stability of the system was poor, as evidenced by the stability score of 55% (Table 3).

According to the NTCP managers, the strengths of the system include the program’s commitment to provide a treatment for cases, good monitoring system, well-trained staff, and strong donor support. On the other hand, the managers stated that dependence on the external fund is a major weakness of the system, especially if stopped. Other weaknesses included unavailability of the MDR-TB center at the governorate level, absent inpatient department, no refresher training or training for new staff, absence of an electronic system, no trained staff for data analysis at the regional MDR-TB centers levels, and no constant electrical power supply for the lab.

All the 16 regional MDR-TB staff agreed that the case definitions are easy and clear. Although 71% reported that the guidelines are easy and understandable, 50% reported that it is preferable to have these guidelines written in Arabic. Regarding data collection, 44% respondents reported that this is difficult because, sometimes, the data are needed to call the patient and 66% reported that the data require regular follow-up. All the respondents agreed that transferring data to the NTCP is easy. However, 56% of the respondents reported that there is shortage in lab tests, especially the routine follow-up lab tests (eg, liver function, renal function, and eye and hearing tests) that are only available at the Sana’a regional MDR-TB center. The majority (94%) stated that they received training only once in 2014, and all stated that they need refresher training. The simplicity attribute has been scored 72%, which means that the system is average in simplicity (Table 4).

All the respondents indicated that they receive quarterly incentives for their work and that they are willing to continue participating in the MDR-TB surveillance system. Overall, 81% of the respondents were satisfied. Acceptability scored 82%, which indicates that the system has good acceptability (Table 5).

By reviewing the regional Sana’a city DR-TB center’s registers and cases files, we found that 5% and 20% had missing data, respectively. However, comparison of data on case files and register did not show any discrepancies.

The performance of the surveillance system is illustrated in Table 6.