Background: We have earlier studied how precipitation effect quality parameters in a fresh water supply for citizens in Gothenburg, Sweden. We concluded that the quantity of preceding rainfall are a superior predictor to concentrations of indicator bacteria's at the river water intake to their drinking water utility. Moreover, we have studied associations between rainfall and nurse advice calls from households that are distributed drinking water produced from this water utility. We identified that heavy rainfall was associated with an increase in daily number of phone calls concerning gastrointestinal illnesses (GI calls) 4-6 days later. We also observed that consecutive days of wet weather increased daily number of GI calls. Aim: Evaluate relations between GI visits to primary health care centers and nurse advice calls in entire, and subparts of, Gothenburg. This is of interests since we are studying causality of disease that normally doesn't need prescribed medicine or a doctor's advice, and this study will enlighten the similarities and differences of these types of health data. Method: We adjusted for seasonality and time trend, day of week and holidays in both GI health indicators separately with use of GAM models. We calculated cross correlations of lags -15:15 of the two model residual series to compare short-term associations. Results: No clear short-term associations were found between visits to primary care centers and nurse advice calls concerning gastrointestinal health (n days=758). Different seasonality patterns were observed with increasing GI visits in late summer compared to GI calls. GI calls have more distinct winter peaks than GI visits. Conclusions: The two different GI incidence data seems to be independent in short-term variations and we suspect that they represent different types of GI illnesses. Persons with virus-like GI symptoms are usually not recommended to visit a health care centers, why nurse advice calls may be a better outcome to study than visits.
展开▼
