Vital Statistics: The problem of exploring the burden of Non-communicable Disease (NCDs) in Sri Lanka, 1950-2000

Posted on 13 June 2017

In the run up to Global Health Histories Seminar 101: 'A world that counts?' Contexts of health: Information for Health’ we are taking a look at the health data issues present in the work of our colleagues on the ‘Local Bases of Global Health’ project. We asked Dr Margaret Jones, a research fellow at CGHH, to reflect on her work on Sri Lanka.

My research on primary health care policies in Sri Lanka in the period 1950-2000 has entailed an exploration of the burden that NCDs have imposed on the health services of the island and its population. The demographic and epidemiological transition that Sri Lanka underwent in this period arising from declining death rates, increasing life expectancy, a decline in infectious disease and an ageing population, brought NCDs to the forefront of the island’s disease burden.

NCDs therefore form a significant part of the primary health care story. But a major obstacle, for both contemporaries and historians, to assessing the significant problem that NCDs presented is the lack of reliable and extensive statistics on the morbidity and mortality from the most common ones – heart disease, cancer and diabetes. Throughout this period the reliability of mortality and morbidity statistics generally cannot be taken for granted but finding information on the prevalence of NCDs presents a particular challenge. The health authorities made little attempt to collect data on NCDs, quite reasonably focusing their constrained resources on the far more pressing problem of infectious diseases.  The mortality and especially the morbidity NCDs caused remained almost entirely undocumented. Extensive statistics on the special campaigns against the infectious diseases – malaria, tuberculosis, leprosy, filiariasis and VD - for example, appear assiduously in annual medical reports, health bulletins and the weekly epidemiological reports issued by the Health Department. But data on NCD morbidity could only be obtained from hospital admission records. Sri Lanka, alongside other developing countries, did not have the resources to collect population-based morbidity statistics.

As awareness of NCDs developed from the 1980s onwards the annual health bulletins pointedly and repeatedly highlighted this knowledge gap, noting that hospital records were the only source of data on morbidity, and that they could therefore present only a partial picture of the burden. There was no data, as for example the 1992 Annual Health Bulletin explained, on those who “were admitted to Ayurveda and private hospitals, and the large numbers who were attending the outpatient departments of Government medical institutions, dispensaries of private practitioners and offices of Consultants”. Furthermore, “the private sector comprising of western and traditional medicine, accounts for 55 per cent of the patient load which is mainly outpatient management”. Added to this there was the factor of self-medication, facilitated at this time as highlighted in this Bulletin by the “laxity of the Cosmetic Devices and Drugs Act which permits the purchase of a wide range of drugs without professional prescription”. The Bulletin concluded therefore that, “the morbidity load, as indicated by statistics obtained from Government hospitals is only the tip of the iceberg”. (Annual Health Bulletin Sri Lanka, Colombo: Ministry of Health, 1992).

The result is that the burden of NCDs was largely invisible for much of the period of my study and it was only as the demographic and epidemiological transition in Sri Lanka took hold that attempts were made to make them visible. But, still determined by the lack of resources, this could only be done through population surveys varying in size, population profile and geographical coverage. These surveys have, however, led one contemporary to talk of an “explosive epidemic of NCDs”(Premaratne, Amarasinghe and Wickremasinghe, “Hospitalisation Trends due to selected non-communicable diseases in Sri Lanka, 2005-2010, Ceylon Medical Journal, 50:2, June 2005). In terms of epidemiology NCDs  are now being given due prominence in the health policy landscape and increasing attention is being given to assessing their true prevalence in order to facilitate the implementation of effective policies to reduce the threat they pose. 


Global Health Histories Seminar 101: 'A world that counts?' Contexts of health: Information for Health’ will be streamed live online on Thursday, the 22nd of June 2017 from Copenhagen (12.30pm CEST or 11.30am GMT).

To learn more about Margaret's research, you can access her staff profile page.