Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

PLOS BLOGS Your Say

Reading between the ‘data’: Reporting behaviours and their implications for COVID-19 documentation

Authors: Rakesh Parashar, Ankita Mukherjee, Samuel Franzen

“Illusion is needed to disguise the emptiness within”- Arthur Erickson

A robust surveillance system of routine health indicators allows policymakers and implementers to make timely evidence-based decisions. If complete and accurate, such data systems can inform the success of interventions and identify priorities by capturing data on service delivery and health outcomes. Within the health ecosystem of India, several data capturing and reporting mechanisms exist- with many using digital platforms. However, no single routine health surveillance system provides accurate and timely data on population health.

COVID-19 reporting suffers from similar issues. In the ongoing COVID-19 pandemic, the highest emphasis of the government dashboard and official media reports has been on daily and cumulative tests done, recovery rates of the total COVID-19 cases, and an absolute number of vaccinations –  often the number of the first doses given. On the contrary, critical data elements such as confirmed positive test cases and covid-19 related deaths are kept under check.

Numerous estimates suggest significant underreporting of the actual number of COVID-19 cases and deaths compared to the reported figures, although the government maintain otherwise. There have been instances where this practice was also called out by the judiciary systems. While these gaps and the need for more accurate data in COVID-19 have been widely discussed, it is seldom inquired why such data discrepancies are common. In this article, we bring forth a systemic perspective on the reasons behind the inconsistencies in the routine health data by drawing on several research studies on health management information systems (HMIS) and our own health program experience in various Indian states.

HMIS is one of the key data management portals that electronically captures the routine public health service delivery data in Indian States. Data from health subcenters and Primary Health Centres (PHCs) are often passed to higher-level facilities in a paper-based format where they are entered into the portal. Given the considerable importance of HMIS data for reporting requirements, the quality of the data entry process is the subject of much interest.

Our study (unpublished) on HMIS-reporting in one of the most populous Indian states underlined the many challenges related to its reporting. Common technocratic constraints included resource availability (data entry personnel, computers/internet) and processes such as difficulty in understanding and filling data formats, among others.

Yet more noteworthy is the politics of reporting and the implications this has for selective reporting or massaging of data so as to present a favourable picture of the health facility or health program performance. This behaviour often determines which indicators get over-and under-reported. Arguably, indicators showcasing health facility, block, or district unfavourably would be attempted to be hidden. In contrast, parameters showing better performance would be overreported and more publicised. We observed that maternal and newborn deaths are frequently underreported by public health facilities, and services like immunisation are generally inflated.

This reporting behaviour is indicative of the top-down authoritative culture prevalent in the local health systems. To gain an appreciation and avoid punitive action for poor performance against targets, the lower-level health managers present data in a way that is perceived as ‘good’ to reflect a false sense of achievement. This practise trickles down all levels of the system, including the most distant rung of health workers -ANMs and ASHAs- who being part of such ‘data management culture’ misreport the services they are responsible for. Entered data may even be further ‘corrected’ at the block and district levels before state-level inspection.

Tangible factors such as acute staff shortage, faulty equipment pieces, and poor data comprehension by staff encourages this behaviour and adversely affects real-time data documentation. The less considered factors include fear of being humiliated for failing to achieve monthly targets, additional paperwork, and use of data to assess staff performance. We found that although staff may wish to report correct data, they are often discouraged from doing so by limited use of quality data for decision-making at facility levels, poor understanding of the significance of data, and lack of time and incentives. These issues are compounded by planning and budgeting processes which set ambitious targets that are unfeasible to achieve without addressing bottlenecks.

Conclusion:

Against this backdrop, it is unsurprising that COVID-19 data are underreported by some Indian states. Despite high quality data being essential for disease control and managing outbreaks, emergency situations frequently exacerbate the problem because of political interest.  Arguably, due to the pandemic being conferred as a national priority, it creates further negative pressure at the grassroots levels, thereby motivating staff to fabricate data even more.

We argue that improving the quality, answerability, and use of data is deeply entrenched in the political nature of systems and reflects societal values. Showcasing high performance despite failures and complex pathways to success trickles down to various health systems and affects the reporting at the frontline. The prevalent culture that disincentivises accurate reporting and favours ‘managed’ data is detrimental to the system’s progress. This obscures the true estimation of critical indicators and negates the sole reason for creating data systems. It ultimately results in flawed decision making when data are used to make decisions around resourcing, accountability, and meeting public demands.

Some studies have observed similar findings in Haryana and Uttar Pradesh, however, there remains limited evidence on the prevalence of these data reporting behaviours in Indian states. While this calls for further attention and research, it is informally accepted to be a common practice.

From what is known, it’s clear that routine approaches such as staff training and the threat of punitive action are unlikely to make a big difference and can be counterproductive. Instilling quality data management is a complex web of systemic, social and interpersonal factors. We hence emphasise investing in cultural reforms in health data reporting, backed by structural reforms such as incentivising accurate and timely data reporting even if the data shows poor performance. The punitive and humiliating culture should be put to rest, and mechanisms to support good quality information should be encouraged.

Pandemic or not, health policy decisions using quality data could help many unsolved problems of the local health systems. It is time that the policy investments are directed towards ‘nurturing’ the local health systems, rather than attempting to fix such issuues superficially.

Author biography:

Rakesh Parashar MBBS PhD, is currently working as health systems and policy expert with Oxford Policy Management. He is a PhD in health systems studies and has a background in public health and medicine.

Ankita Mukherjee MSc MPH, is presently working as a Research Assistant at Oxford Policy Management. She has interests in maternal and newborn health.

Dr Samuel Franzen is an International Development and Project Management professional specialising in monitoring evaluation research and learning. He has over 15 years’ experience working in the field of health systems strengthening, quality/service improvement, risk assessment and quality assurance, research uptake and communication.

Leave a Reply

Your email address will not be published. Required fields are marked *


Add your ORCID here. (e.g. 0000-0002-7299-680X)

Related Posts
Back to top