Management of childhood diarrhoea at the household level: a population-based survey in north-east Brazil.

The management of childhood diarrhoea at the household level was studied in a population-based survey in four states in north-east Brazil. Of a representative sample of 6524 children under 5 years of age, 982 (15.1%) had diarrhoea on the day of the interview or had had diarrhoea at some time during the previous 15 days. A total of 66% of the children were not taken for treatment, while government health services were used by 14%, private doctors by 1%, and traditional healers (rezadeiras) by 24%. Oral rehydration therapy was given to 24.3% of the children as follows: solutions of oral rehydration salts (ORS) were received by 6.8%, salt-and-sugar solutions by 14.7%, and solutions of commercial ORS brands by 4.3%. Although 95% of the caretakers knew about rehydration solutions, only 18% prepared them correctly, the most common error being the use of insufficient water. Of the rehydration solutions used, 39% had a sodium concentration that was potentially dangerous (greater than 120 mmol/l), and 8% had a sodium concentration that was very low. Of those solutions prepared using ORS, 38% had too high a sodium concentration, while 14% of the salt-and-sugar solutions prepared using either the "scoop-and-pinch" approach or a plastic spoon were too concentrated. However, potentially the most dangerous were the salt-and-sugar solutions prepared using nonstandard recipes. More than half of these had an unacceptably high sodium concentration or osmolarity.

A comparison of three methods of measuring maximal opening of the mouth.

Two extraoral techniques are compared with a direct technique for measuring maximal opening of the mouth. The extraoral techniques allow recording of the variable independent of the teeth. The accuracy and precision of the three techniques are calculated and compared statistically, and the significance of the results are discussed.

Studies with pain rating scales.

Good correlation has been shown between pain scores derived from 4 different rating scales. The correlation was maintained when presentation of the scales was separated by a series of questions and by physical examination. There is good evidence that the 4 scales are measuring the same underlying pain variable as they calibrate well. There is also evidence that an 11-point (0-10) numerical rating scale performs better than both a 4-point simple descriptive scale or a continuous (visual analogue) scale.