ing self-referrals to the emergency department in Alabama did not increase.
The projected incremental costs to Louisiana during the discontinuance of excess health care facilities visits were estimated at $1.4 million. This can be compared with the $400,000 in savings to the state from closing its center. In short, had the state spent the $400,000 to keep its poison control center open, it would have saved the system $1.4 million, for a net savings of $1 million. Phrased differently, this implies a savings to the health care system of more than $3 for every dollar invested in the center. This is an underestimate of the benefits of the poison control center insofar as it does not take into account a reduction in mortality and morbidity, or in anxiety and time to the public.
Phillips et al. (1998) used the results of another “natural experiment” to examine cost savings of poison control centers. Between 1993 and 1994, a single county in California lost funding for its center. Public callers to the center received a recorded announcement advising them to dial 911 for poisoning exposures and information. If they called 911, they were patched into a neighboring poison control center to which they had prior direct access. An analysis was done of individuals who called the center during this interruption of service compared with a matched set who called subsequently after service was resumed. The outcomes during the period of blockage were substantially different than during the control period, even though the disruption did not involve lack of access to the center, but only patching into one through 911. Fourteen percent of callers with restricted access were treated in an inappropriate location (e.g., treated by an emergency department when they might have been managed at home), compared with 2 percent who had direct access to a poison control center. In a further analysis of the costs associated with the same blocked-caller episode, it was found that restricting access resulted in an additional $10.98 per case in net societal costs (all costs and benefits, including patient time and transportation and marginal costs for resources used as a result of the block) and an additional $33.14 per case in health care purchaser costs (Olson et al., 1999).
Harrison et al. (1996), in one of the most thorough of the existing analyses, adopted a decision theoretic analysis to evaluate treatment management guidance for the public. In addition to secondary data on costs such as emergency department visits, ambulances, and other factors, Harrison and colleagues used data assembled from an expert panel of toxicologists to estimate probabilities of morbidity outcomes, mortality outcomes, and adverse treatment impacts of cases coming into an emergency department. Thus the researchers were able to consider not only differences in direct costs to the health care system, but also differences in morbidity and mortality due to the provision of poison control center