Human touch to detect hypothermia in neonates in Indian slum dwellings.

OBJECTIVE To assess the validity of human touch (HT) method to measure hypothermia compared against axillary digital thermometry (ADT) and study association of hypothermia with poor suckle and underweight status in newborns and environmental temperature in 11 slums of Indore city, India. METHODS Field supervisors of slum-based health volunteers measured body temperature of 152 newborns by HT and ADT, observed suckling and weighed newborns. Underweight status was determined using WHO growth standards. RESULTS Hypothermia prevalence (axillary temperature <36.5 degrees C) was 30.9%. Prevalence varied by season but insignificantly. Hypothermia was insignificantly associated with poor suckle (31% vs 19.7%, p=0.21) and undernutrition (33.3% vs 25.3%, p=0.4). HT had moderate diagnostic accuracy when compared with ADT (kappa: 0.38, sensitivity: 74.5%, specificity: 68.5%). CONCLUSIONS HT emerged simpler and programmatically feasible. There is a need to examine whether trained and supervised community-based health workers and mothers can use HT accurately to identify and manage hypothermia and other simple signs of newborn illness using minimal algorithm at home and more confidently refer such newborns to proximal facilities linked to the program to ensure prompt management of illness.