Quality, cost, and outcome of intensive care in a public hospital in Bombay, India

Objective: To study the quality, cost, and benefits of intensive care in a public hospital in Bombay, India. Design: Prospective collection of data. Setting: Seventeen-bed medical-neurology-neurosurgery intensive care unit ( ICU) of a municipal teaching hospital. Patients: A total of 993 consecutive ICU patients during a 16-month period. Interventions: None. Measurements and Main Results: The 993 patients aged 36.5 +/-. 16 yrs (mean +/- SD) had a day-1 Acute Physiology and Chronic Health Evaluation (APACHE ) II score of 14.9 +/- 9,6 (mean +/- SD), with a predicted mortality of 21. 7%; the observed mortality was 36.2% (standardized mortality ratio = 1.67). The day-1 Therapeutic Intervention Scoring System (TISS) points were 17.7 +/- 6.2 (mean +/- SD), and total TISS points per patient were 87.6 +/- 110 (mean +/- SD). Nurse-to-patient ratio in the ICU was 3:17 and the average w orkload per nurse was 64.2 TISS points. The average length of stay was 5.5 days (SD = 7.1 days). The overall cost of treating 993 patients was, in Ind ian rupees (Rs), Rs 107,79,209 (U.S. $307,997), and cost per patient per da y was Rs 1,973 (U.S. $57). The cost per survivor was Rs 17,029 (U,S, $487) and cost per TISS point was Rs 90.14 (U.S, $2.57), The low cost per TISS po int was attributable to the reuse of disposable equipment and lower cost of drugs and salaries for medical and paramedical staff. Conclusions: Intensive care in India is cheaper than in the West; however, mortality is 1.67 times that for patients with similar APACHE II scores in ICUs in the United States. This finding may be attributable to the lesser i ntensity of care per patient (lower day-1 TISS points), lower nurse-to-pati ent ratio because of shortage of trained personnel and budgetary constraint s, and higher workload per nurse (64.2 TISS points per nurse, compared with 40 points per nurse in the West). In addition, the APACHE II scores may un derestimate mortality for Indian patients because of differences in case mi x, higher lead time between onset of admission and treatment before ICU adm ission, and possible inappropriateness of age points derived from American patients for Indian subjects because of a higher burden of diseases at lowe r ages in Indian patients.