Unmet need and met unneed in health care utilisation in Iran

Purpose - This paper uses a unique nationwide survey data derived from the 2003 Utilisation of Health Services Survey (UHSS) in Iran (n=16,935) to analyse inequities of health care utilisation. Design/methodology/approach - Concentration indices are used to measure socioeconomic inequality in actual use of the five types of health services, and in unmet need for two of those types of service (any ambulatory care and hospital admissions). Horizontal inequity indices are employed to examine inequity in ambulatory and hospital care. Generalised linear model (GLM) was employed to investigate factors contributing to the phenomena of “unmet need” and “met unneed”. Moreover, a decomposition analysis of inequality is performed to determine the contributions of each factor to the inequality of “unmet need”. Findings - Results suggest that self‐reported need for ambulatory and inpatient care is concentrated among the poor, whereas the utilisation of ambulatory and inpatient care were generally distributed proportionally. Results of horizontal inequity indices show that the distributions of any ambulatory care and hospital admissions are pro‐rich. The probability of “unmet need” for ambulatory care was higher among wealthier individuals. The decomposition analysis demonstrates that the wealth index, health insurance, and region of residence are the most important factors contributing to the concentration of “unmet need” for ambulatory health care among the poor. Results also illustrate that higher wealth quintiles used more unneeded ambulatory care than their poorer counterparts. Originality/value - A special characteristic of the UHSS is that it contains questions about the need for medical services use and about actual services use. This characteristic provides an opportunity to measure the inequality of health care consumption against self‐assessed treatment needs, as well as an analysis of which observables are associated with “unmet need”. Moreover, the incidence of health care use when it is reported as not needed can be analysed with this dataset. The analysis of this phenomenon – which we refer to as “met unneed” – is another novel aspect of this work.