Pulmonary tuberculosis and deprivation in hospitalised patients in Scotland C. Hadjichristodoulou, P. Christie & S. O'Brien Scottish Centre for Infection and Environmental Health, Ruchill Hospital, Glasgow G20 9NB, Scotland Accepted in revised form 11 April 2001 Abstract. During the last decades tuberculosis re- emerged in almost all over the world, in both de- veloped and developing countries. Many risk factors were implicated to explain the re-emergences in- cluding the HIV epidemic. The aim of the study was to explore if tuberculosis is related with poverty in Scotland utilising routinely collected hospital dis- charge data for patients with pulmonary tuberculosis and postcode-derived Carstairs deprivation scores. The Carstairs and Morris index is composed of four indicators which were judged to represent material disadvantage in the population. A positive correla- tion was found between the cumulative incidence rate for hospitalised patients within each Health Board and the Carstairs deprivation score (r  0:76, p < 0:01). A similar correlation was found between the cumulative incidence rate and the deprivation scores within each postcode sector (r  0:47, p < 0:0001). These results supports ®ndings by other researchers that poverty and tuberculosis are related, and might be one explanation for the recent re- emergence of tuberculosis. Key words: Tuberculosis, Deprivation, Re-emergences of tuberculosis Introduction After the establishment of eective chemotherapy for tuberculosis in 1950 an acceleration of the rate of decline of noti®cations from tuberculosis was re- corded in Scotland. The decline of noti®cation rates for tuberculosis continued until the mid-1980s when the annual number of noti®cation appeared to pla- teau, a situation, which persisted up to 1996 when a small decline was again recorded as shown in Fig- ure 1. Similar situation was observed in other Euro- pean countries including England and Wales. In some European countries (e.g. Switzerland and Holland), an increase in total number of noti®cations of tu- berculosis re¯ects an increasing contribution from immigrant and refugee populations in these nations. In some countries a signi®cant contribution to annual tuberculosis noti®cations is made by individuals with HIV infection. In Scotland, the contribution of HIV infection and ethnic minorities to arresting the decline of tuberculosis is negligible [1]. Scotland has low immigrant population (less than 1% of population are from the Indian sub-continent), few refugees and HIV co-infection contributes less than 2% to tuber- culosis noti®cations in Scotland. Tuberculosis has always been related to poverty, and previous studies have reinforced this observation for England and Wales [2, 3]. The objective of the current study was to explore if a similar relation- ship exists in Scotland utilising routinely collected hospital discharge data for patients with pulmonary tuberculosis (PTB) and postcode-derived deprivation scores. Methods Every Scottish hospital discharge generates a detailed record (the SMR1 form) which includes up to six di- agnosis ®elds (ICD9 coded) in addition to hospital, specialty and health board identi®ers and postcode of residence. SMR1 records containing ICD9 codes for PTB(011.0±011.9)inanyofthesixdiagnosis®eldsand covering 5 year period (1990±1994) in which the pla- teau of noti®cations was recorded, were obtained from the Information and Statistics Division of the Com- mon Services Agency. Decoding software was used to select 2504 discharge records in which PTB was in- cludedasthemain(®rst)diagnosis(2420discharges)or as a second diagnosis to a relevant symptom in ®rst position (84 discharges). These 2504 discharges relate to 1773 patients, some patients having had multiple admissions or transfers between clinical specialties. The 1773 patients were included in the analysis to- gether with the Carstairs deprivation scores. The CarstairsandMorrisindexwasoriginallydevelopedin the 1980s using 1981 census data. It is composed of four indicators which were judged to represent mate- rial disadvantage in the population. The four indica- tors are combined to create a composite score. The deprivation score is divided into seven separate cate- gories,rangingfromveryhightoverylowdeprivation. The seven categories were designed so as to retain the discriminatory features of the distribution of the deprivation score, rather than to ensure equality of numbers between each deprivation category. The variables used to calculate the scores were as fol- lows: European Journal of Epidemiology 17: 85±87, 2001. Ó 2001 Kluwer Academic Publishers. Printed in the Netherlands.