The Spanish TB national surveillance system is passive, case-based, and at the population level; each CCAA is responsible for collecting and annually reporting patient-level data on predefined variables into the Spanish Surveillance System (SiViEs Plus) according to the TB protocol published by The National Epidemiological Surveillance Network (La Red Española de Vigilancia Epidemiológica [RENAVE]; Figure 1) [10].

We extracted data on 40 key variables from all reported cases in Spain from the SiViEs Plus database between January 1, 2018, and December 31, 2023. We divided variables into 3 categories: (1) patient-related information (11 variables), (2) illness-related information (18 variables), and (3) laboratory-related information (11 variables).

For the second aim, we developed a questionnaire based on a literature review and consulting with TB and surveillance experts at the CNE. The 17-question online questionnaire included both closed and open-ended questions to gather information on the perceived usefulness of the annual feedback reports, challenges around TB surveillance, and suggestions on how to improve TB surveillance in Spain. We piloted the questionnaire with the TB surveillance lead from 1 CCAA. In April 2025 and May 2025, we circulated the final version via Google Forms among the Spanish TB Surveillance Working Group. The survey was completed by the lead individual responsible for TB surveillance in each CCAA; these individuals were considered best placed to respond on behalf of their CCAA as they have an overview of all areas of TB surveillance in their region, they are responsible for submission of data at the national level, and they are the person who receives the annual feedback reports. The survey invitation emphasized that results would be reported anonymously and not attributable to any CCAA or individual.

This study used anonymized data from RENAVE in Spain. RENAVE collects data on cases notified through the national reporting electronic platform SiViEs Plus, and it is hosted by the CNE. These data are generated under a legal public health mandate and are routinely used for epidemiological monitoring. As the study involved secondary analysis of anonymized administrative surveillance data accessed under legally mandated confidentiality safeguards, formal ethics committee approval was not required. This exemption is consistent with Spanish and European data protection regulations governing the processing of anonymized public health surveillance data, including Ley Orgánica 3/2018 de Protección de Datos Personales y Garantía de los Derechos Digitales (LOPDGDD) and Reglamento (UE) 2016/679 (GDPR).

Survey data were submitted by a regional focal point from CCAAs in Spain. For survey participation, informed consent was obtained from participants prior to beginning the survey. Participants were made aware that taking part was voluntary, that there would be no consequences for not taking part, and that they could withdraw from participating at any time without consequence. One response was obtained from each CCAA; however, responses were reported anonymously in the analysis and write-up to maintain confidentiality. All data were stored and analyzed within secure, password-protected ISCIII systems in compliance with institutional data protection policies.

Internal completeness and acceptability surveillance attributes were chosen as they align with those set out by the European Center for Disease Control and Prevention (ECDC) technical handbook “Data quality monitoring and surveillance system evaluation” [11] as key for high-functioning surveillance systems.

We defined internal completeness according to ECDC [11] as “the number of completed data fields out of the total number of data fields (unknown and missing items should be included in the denominator).” The denominator was not always equal to the total number of cases; for example, the variable “number of years living in Spain” was only applicable to cases born outside of Spain.

TB data are submitted continuously by CCAAs to SiViEs Plus; however, each annual dataset is only considered complete once regions confirm that all records for the previous year have been uploaded, typically in Q2 of the following year (eg, annual data for 2021 would be consolidated in Q2 of 2022).

Since 2022, following the annual submissions from the CCAAs, the national level (CNE) performs data validation and generates feedback reports highlighting missing or inconsistent information (see Multimedia Appendix 1 for an example report in the original Spanish and in an English translation that was carried out using the generative artificial intelligence [AI] tool, Copilot). These reports are typically shared with CCAAs between May and July, with variation in timing depending on when each region finalizes its initial submission. CCAAs subsequently update their annual data before final national consolidation, which usually occurs in the third quarter (Q3). The feedback reports disseminated in a given year are therefore designed to inform and improve data reporting practices for the year prior. As feedback reports began being disseminated in 2022, the year 2021 was considered the first year in which reporting could be influenced by the feedback process.

TB surveillance data for 25,229 reported cases were extracted across the study period. In period 1 (2018‐2020), there was a total of 13,328 cases, and in period 2 (2021‐2023), there was a total of 11,901 cases. Comparing internal completeness of 40 variables across period 1 and period 2, there was a mean improvement of 10.2% (Wilcoxon signed rank test P<.001), from 66% in period 1 to 76.2% in period 2, with 78% (31/40) variables showing an increase (Table 1). No change was observed for 4 (10%) of the 40 variables (age, sex, TB classification, date of case report), which all had a completeness rates ≥99.9% in period 1, while 5 (13%) of the 40 variables had a decrease in completeness (country, CCAA, province and municipality of residence, and mother number of years in Spain; Table 1).

Patient-related variables had the highest overall completeness for both periods (80.3% for period 1 and 82.7% for period 2). However, the mean increases of 2.4% across the 11 variables was lower than that of other variable groups and was not statistically significant (Wilcoxon signed rank test P=.48). All 5 variables with a decrease in completeness were within this category.

Illness-related variables significantly increased by 12.8% (paired t test: t₁₇=–5.185, P<.001), from 64.6% in period 1 to 77.4% in period 2. Of the 18 variables within this group, 2 did not increase (TB classification and date of case report) as they were already at 100% completion during period 1.

The greatest improvement was seen with laboratory-related variables, with an increase of 13.9% (paired t test: t₁₀=–4.392, P=.001) from 53.9% in period 1 to 67.7% in period 2. All 11 variables increased, the largest of which was for “date of laboratory diagnosis,” which increased by 37.5%.

There was a large degree of variation across the 17 CCAAs and the 2 autonomous cities (Figure 2). For period 1, mean annual percentage completeness across all variables ranged from 44.9% to 88.7%, and for period 2, it ranged between 56.9% and 87.5%. Comparing period 1 with period 2, differences in completeness ranged from a decrease of 5.2% to an increase of 26.2% in period 2. For all regions, 84% (16/19) had an improvement in period 2, 69% (11/16) of which had statistically significant improvements (P<.05). Of the 3 (3/19, 16%) CCAAs where a decrease in completeness was observed, only 1 was statistically significant.

Results obtained were similar across all analyses when the pandemic years of 2020 and 2021 were excluded (Table S2 in Multimedia Appendix 2). Across all 40 variables, there was a statistically significant increase of 12.5% (P<.001), compared with 10.2% (P<.001) in the original analysis. For patient variables, results were identical to the original analysis, with a nonstatistically significant increase of 2.4% observed. Differences between the 2 periods were slightly greater for illness-related variables and laboratory-related variables when excluding the pandemic years: 14.8% (P<.001) compared with 12.8% (P<.001) and 18.9% (P<.001) compared with 13.9% (P=.001), respectively.

In June 2025, to support CCAAs with making continuous improvements in TB surveillance, we shared with regions a summary of Part 1 and Part 2 of the assessment, along with a description of changes made to the upcoming feedback reports (for 2024 data submission) based on the findings:

Following this, we analyzed the preliminary 2024 TB surveillance data to determine whether the updated feedback report supported further improvements, as part of a continuous improvement cycle. We found that, at the national level, mean completeness across all variables increased to 78% in 2024 (compared with 66% in 2018‐2020 and 76% in 2021‐2023).

This study aimed to quantitively assess the internal completeness of TB surveillance data before and after implementation of feedback reports in Spain and to qualitatively assess the perceived usefulness among CCAAs of both the current surveillance system and the feedback reports. We identified significant improvements in completeness across most variables and regions following the introduction of annual feedback reports, although there was some variation across both type of variable and CCAAs. At the national level, mean completeness increased from 66% in 2018‐2020 to 76.2% in 2021‐2023 (with nearly four-fifths of variables indicating improvements) and increased further to 77.8% in 2024. Both the surveillance system and feedback reports were generally viewed as acceptable and useful among regional stakeholders included in the survey. The additional study aiming to update the feedback reports for the 2024 surveillance cycle based on our findings and assess the potential impact of the updated reports on TB surveillance data completeness found that mean completeness continued to improve in 2024 relative to previous years.

Although the highest completeness levels were seen among patient-related variables, no statistically significant improvement was observed in this group. This is partly due to many variables being at or close to 100% completeness in period 1, leaving minimal opportunity for further improvement. This is in line with findings from other European Union (EU) and European Economic Area (EEA) countries, where, in 2023, most countries had complete or near complete data (>95%) for variables requiring mandatory reporting to the ECDC, such as age and sex [12]. Patient-related variables that continued to have low completeness rates in period 2 were those that applied to individuals born outside of Spain: number of years in Spain, mother country of birth, and mother number of years in Spain (62.3%, 12.3%, and 55.3% period 2 completeness, respectively). This is pertinent, as individuals from countries of origin with higher TB prevalences may have an increased risk of importation; therefore, efforts should be made to improve data collection for these variables.

Although illness-related variables had similar percentage completeness (77.4%) to that of patient-related variables in period 2, a larger improvement of 12.8% in illness-related variables was observed relative to period 1. However, some variables remained poorly completed, including follow-up result, which, despite a 23.8% increase, had 39.3% completeness in period 2. Supporting this, survey respondents called for improved protocols and better integration of case follow-up and laboratory information, particularly for patients who move between CCAAs. The respondents’ call is in the context of the WHO recommendations for a 6-month regimen of 4 first-line drugs (and longer treatment and additional drugs for resistant TB) [13], which complicates patient follow-up. Continuity of TB treatment is essential for effective control of the disease, particularly with increasing movement of individuals both within and between countries across Europe [14]. In particular, interruption of treatment among mobile populations (including migrants, refugees, and asylum seekers) is a recognized challenge in Europe, often leading to poorer health outcomes, higher risk of transmission, and the emergence of resistant strains [15]. In line with this study, evidence from other countries has found that variability regarding migrant health policies and unclear or absent referral mechanisms can hinder treatment continuity [6,14,16].

Improvements were most pronounced for laboratory-related variables, with a mean increase of 13.9%; however, increases were particularly pronounced for those with lower baseline completeness, such as for date of laboratory diagnosis and variables related to antimicrobial susceptibility testing. Although this highlights substantial improvements in capturing microbiological information, challenges persist due to the lack of interoperability between clinical and laboratory systems. Despite improvements, some laboratory variables continue to have low completeness in period 2, including lab tests carried out (50.6%) and results for the 4 included antimicrobial resistance testing variables, which ranged from 31.9% to 50.4%. This finding reinforces the need to consolidate mechanisms for the automatic integration of microbiological data into epidemiological surveillance, as demonstrated effectively in some CCAAs. In Catalonia, for example, which ranks higher than most CCAAs with respect to completeness in period 1 (7th) and period 2 (6th), laboratory results are linked to the regional surveillance system [17]. At the European level, models such as the MiBa (Danish Microbiology Database) in Denmark [18], Osiris system in the Netherlands [19], and MycobNet network in the United Kingdom [20] enable automated laboratory result reporting and the genomic integration of Mycobacterium tuberculosis strains, improving outbreak detection and resistance surveillance.

This finding that variables requiring integration between clinical and laboratory systems or ongoing follow-up were the most incomplete is consistent with the ECDC [11] and United States Centers for Disease Control and Prevention (CDC) guidance [21], emphasizing that data involving multiple sources or complex definitions are generally more vulnerable to missingness. In line with this, across EU/EEA countries in 2023, the availability of data related to clinical and laboratory information was lower than for patient-related variables, for example for previous treatment history (86%), treatment outcome at 12 months (61.5%), and HIV status (39.3%) [12].

At the regional level, 16 of Spain’s 19 CCAAs (84%) saw an improvement in period 2, although the magnitude varied widely, ranging from modest gains to more than 25% increases. The one region where a statistically significant decrease was observed already had a high level of completeness in period 1 (the 5th highest among all CCAAs). These disparities reflect the decentralized nature of the Spanish surveillance network RENAVE. This is in line with other countries with decentralized surveillance systems, where regional disparities have been found in evaluations of the surveillance systems, for example in Italy where northern regions were found to report acute viral hepatitis cases in a more timely manner than central and southern regions [22].

Survey findings provided important explanatory context. Respondents were generally satisfied with TB surveillance in Spain and considered it to be important, indicating high level of acceptability and perceived usefulness of the system, providing a strong foundation for future quality improvement initiatives. There appeared to be a possible association between higher-performing CCAAs and the extent to which the feedback reports are disseminated among colleagues within the region; further investigation would be required to understand whether this is because the reports are more effective at driving improvements if disseminated more widely or simply that higher-performing CCAAs have both better communication channels and better surveillance systems.

Overall, our findings show an improvement in internal completeness of TB surveillance data following implementation of feedback reports and positive perceptions of these reports among regional stakeholders. Although a causal link is difficult to establish, our findings suggest that systematic data quality monitoring and feedback may help to strength surveillance, consistent with international guidance that emphasizes regular evaluation as a cornerstone of high-performing systems [11,21]. The observed improvements, together with respondents’ endorsement of feedback reports, suggest that feedback cycles may contribute to improved completeness and system acceptability and continuous quality improvement.

This analysis is subject to several limitations. First, due to the descriptive nature of the analysis, it was difficult to establish a causal link between the observed improvements in data quality and the implementation of the feedback reports. For this reason, we included the qualitative element of the survey, but it is not possible to rule out other explanations for the observed results, such as improvements in surveillance across all areas or increased focus on infectious diseases after the COVID-19 pandemic. Furthermore, although improved completeness likely reflects better data capture, it does not guarantee data accuracy; although the feedback reports were designed to examine this by including validity cross-checks, there are variables for which this was not possible. Case ascertainment and the sensitivity of the system—the number of cases captured by the system divided by the true number of cases in the community [11]—were flagged as a potential area of risk by survey respondents and remains an important attribute of surveillance that has yet to be explored.

There were additional limitations related to the survey component. The survey was circulated to the lead individual responsible for TB surveillance in each CCAA, meaning that there was only one response per region. This individual is the person receiving the feedback reports so they were considered best placed to respond to our survey; however, it is possible that this could introduce bias and may not fully reflect the breadth of opinions of individuals in the CCAA. In addition, although it was emphasized that survey responses were anonymous, there may remain some degree of social desirability bias or underrepresentation of more critical perspectives.

Comparisons with other European countries with similar decentralized systems remains a challenge given the limited number of similar studies published internationally. It is hoped that this study will serve as a useful tool for other countries wishing to carry out future assessments examining the attributes of internal completeness, acceptability, and usefulness.

Surveillance of TB is inherently challenging given the typically long-term course of the disease and treatment within the context of complex socioeconomic risk factors. This study indicated that completeness of TB surveillance data in Spain has improved following the introduction of annual feedback reports in 2022, although variation persists by variable and region. Importantly, the combination of quantitative and qualitative analyses highlights not only the measurable progress in surveillance quality but also the perceptions, barriers, and enablers that shape the system’s performance. Although the reasons underlying observed improvements can be uncertain and are likely multifaceted, feedback reports were valued by regional stakeholders and appear to support a culture of continuous quality improvement.

Our findings suggest that the CNE should continue to send annual feedback reports to regions; however, they could be refined to be of even more benefit, for example, by focusing on a smaller number of key variables and indicators that align with those set out by the WHO guidance on TB surveillance published in 2024 [23]. This guidance defines a set of core indicators for countries with a case-based digital surveillance system as well as recommended reporting intervals, visualization, and interpretation for these indicators. This would support regions to target efforts where they are most effective and facilitate comparison with other countries. More broadly, strengthening alignment with international standards may help Spain position itself within global TB monitoring frameworks and contribute more effectively to regional benchmarking efforts.

In addition, the CNE should provide improved guidance on follow up of individuals who move between CCAAs or internationally during the long treatment course and how information from these individuals should be reported in the surveillance system. This could be supported by encouraging use of the ECDC “International tuberculosis care transfer form” [24]. This tool aims to support cross-border collaboration between health care providers to ensure the care needs for people with TB are transferred securely and efficiently, minimizing delays and reducing the risk of treatment interruption. In addition, the availability of a national personal identifier in Spain would facilitate interoperability of laboratory and surveillance systems within regions, as well as support easier transfer of care of patients between regions. These improvements would not only enhance data quality but also strengthen continuity of care, an essential component of TB control in decentralized health systems.

To fully realize the potential of TB surveillance, future assessments could consider expanding beyond completeness and perceptions of feedback reports, for example using standardized methods to carry out a formal evaluation and examining a more comprehensive list of attributes, such as validity, timeliness, representativeness, and sensitivity, in line with established frameworks [11,21,25]. Future evaluations could be also carried out using the WHO TB surveillance guidelines [23] as a framework, with completeness of data fields measured against those required to calculate the recommended indicators. In addition, to examine the possible determinants of changes observed beyond descriptive analysis, approaches such as multilevel logistic regression analysis with adjustment for covariates including CCAA could be considered to complement future evaluations. Finally, embedding assessments into a continuous improvement cycle, as already initiated by revising the most recent feedback reports, will help ensure that the system evolves and continues to provide reliable and actionable information that meet both national and international TB control objectives.