Open Access Highly Accessed Research article

Monitoring the prevalence of chronic conditions: which data should we use?

Juan F Orueta12*, Roberto Nuño-Solinis2, Maider Mateos2, Itziar Vergara345, Gonzalo Grandes6 and Santiago Esnaola7

Author Affiliations

1 Osakidetza, Basque Health Service, C/ Alava n° 45, Vitoria-Gazteiz 01006, Spain

2 O+berri, Basque Institute for Healthcare Innovation, Plaza Asua 1, Sondika 48150, Spain

3 Primary Care Research Unit-Gipuzkoa, Osakidetza, P. Dr Beguiristain s/n, Instituto Biodonostia, San Sebastian, 20014, Spain

4 Red de investigación en servicios de salud en enfermedades crónicas (REDISSEC), P. Dr Beguiristain s/n, Instituto Biodonostia, San Sebastian, 20014, Spain

5 Asociación Centro de Excelencia Internacional en Investigación sobre Cronicidad. Kronikgune, P. Dr Beguiristain s/n, Instituto Biodonostia, San Sebastian, 20014, Spain

6 Primary Care Research Unit-Bizkaia, Osakidetza, c/Luis Power 18 - 4ª Planta, Bilbao, 48014, Spain

7 Department of Health and Consumer Affairs of the Basque Country, C/ Donostia-San Sebastián 1, Vitoria-Gasteiz, 01010, Spain

For all author emails, please log on.

BMC Health Services Research 2012, 12:365  doi:10.1186/1472-6963-12-365

Published: 22 October 2012



Chronic diseases are an increasing threat to people’s health and to the sustainability of health organisations. Despite the need for routine monitoring systems to assess the impact of chronicity in the population and its evolution over time, currently no single source of information has been identified as suitable for this purpose. Our objective was to describe the prevalence of various chronic conditions estimated using routine data recorded by health professionals: diagnoses on hospital discharge abstracts, and primary care prescriptions and diagnoses.


The ICD-9-CM codes for diagnoses and Anatomical Therapeutic Chemical (ATC) codes for prescriptions were collected for all patients in the Basque Country over 14 years of age (n=1,964,337) for a 12-month period. We employed a range of different inputs: hospital diagnoses, primary care diagnoses, primary care prescriptions and combinations thereof. Data were collapsed into the morbidity groups specified by the Johns Hopkins Adjusted Clinical Groups (ACGs) Case-Mix System. We estimated the prevalence of 12 chronic conditions, comparing the results obtained using the different data sources with each other and also with those of the Basque Health Interview Survey (ESCAV). Using the different combinations of inputs, Standardized Morbidity Ratios (SMRs) for the considered diseases were calculated for the list of patients of each general practitioner. The variances of the SMRs were used as a measure of the dispersion of the data and were compared using the Brown-Forsythe test.


The prevalences calculated using prescription data were higher than those obtained from diagnoses and those from the ESCAV, with two exceptions: malignant neoplasm and migraine. The variances of the SMRs obtained from the combination of all the data sources (hospital diagnoses, and primary care prescriptions and diagnoses) were significantly lower than those using only diagnoses.


The estimated prevalence of chronic diseases varies considerably depending of the source(s) of information used. Given that administrative databases compile data registered for other purposes, the estimations obtained must be considered with caution. In a context of increasingly widespread computerisation of patient medical records, the complementary use of a range of sources may be a feasible option for the routine monitoring of the prevalence of chronic diseases.

Chronic disease; Prevalence; Information systems; Computerized medical record systems; Health care surveys; Clinical coding