A time series study of anti-diarrheal drug sales and tap-water quality

Abstract
The City of Le Havre in France operates two water treatment plants and distributes water to 200,000 people. It exploits, as the source of its drinking water, karstic resources that are subjected to episodical microbiological quality degradations. The first plant only chlorinates the water, whereas the second plant normally uses direct sand filtration before chlorination but can also implement coagulation-settling when turbidity of the raw water exceeds 3 nephelometric turbidity units (NTU). During the study period there were several occurrences of failure to maintain residual chlorine as well as significant turbidity increases. The treated water still met all microbiological criteria for potable water in France and this study was undertaken to determine if public health was adequately protected. An ecological time series study was carried out on data collected between April 1993 and September 1996. The record of the sales of medications prescribed or self-selected for the treatment of gastroenteritis (GE) were provided by the pharmacists participating in the epidemiological surveillance network of Le Havre. Sales data, residual chlorine and turbidity measurements were analyzed. Interruption of chlorination of the unfiltered water resulted in a significant increase of medication sales 3 to 8 days later. Raw water turbidity increases resulted in increases of medication sales during the 3 following weeks. The data analyzed suggest that about 10% of the annual cases of gastrointestinal illnesses could be due to the consumption of tap water. This annual average does not reflect the proportional attributable risk occurring during specific periods and underscores the fact that current regulations still do not provide complete protection of public health. Furthermore, such failures have the potential of causing major GE outbreaks if raw water microbiological quality degrades significantly after rainfall events.