Program planning is often myopic and typically underestimate the costs of service innovations.
A 2013 article in the New York Times by Ellen Creswell (http://www.nytimes.com/2013/12/26/health/er-costs-for-mentally-ill-soar-and-hospitals-seek-better-way.html?_r=0&pagewanted=print) describes the problem of finding “a way to help mentally ill patients without admitting them to emergency rooms, where the cost of treatment is high – and unnecessary” and then goes on to describe a solution being explored by a North Carolina medical center. The solution involves paramedics who perform mental health exams on patients in the field, evaluate their mental conditions and then, if persons’ conditions permit, offer options of “being taken to another facility that might be better suited to provide the kind of care they need.” The solution sounds a little like Assertive Community Treatment (ACT). So, might a solution like this lower the cost of emergency treatment?
Our mental health planning model, PBN Free (http://model.planningbythenumbers.org/) can give us an idea of how the above solution might work by comparing two scenarios, one like the current NC mental health system and one like the proposed “solution.” The answer is not as obvious as you might expect; but it’s fairly simple arithmetic. Let’s assume a mental health system provides services to 100,000 people, distributed by functional level in a typical way. One scenario (Scenario 1) is that a basic set of services is provided and that this set is like the one provided by the current North Carolina service system – that is lacking in very many types of community living and support services. In this scenario emergency room usage will be high and seemingly costly, but this will be mitigated by the fact that disappearance rates are high, keeping the number of persons to be served relatively constant or causing it to decrease. Additionally, system costs will be lower since few alternative services will be provided. Another scenario (Scenario 2) is one in which more needed services are provided as the North Carolina solution seems to propose. In this scenario PBN Free suggests emergency service utilization and costs will become a smaller proportion of mental health system costs; but the absolute amounts of emergency service utilization and costs might actually be higher because more persons are maintained in the mental health system. In this scenario, total system costs might also higher because more different types of service are provided.
Tables 1 and 2 and Chart 1, below, show the disappearances and expenditures by service type for the basic and
To get a sense of what the actual impacts of enhancing services in a manner that reduces emergency room costs consider Table3 based on data from Gilmer et al. (2010). This study compared outcomes and costs for 209 persons with serious mental illness (SMI) participating in a full service participation program (FSP) in San Diego County, California. The FSP includes services like Housing First (which is often accompanied by ACT) . For a fuller description of full-service participation programs in California see Gilmer et al. (2010). This table shows that by providing a full package of community support services emergency room costs can be reduced by $1,305. Inpatient and justice system costs are also reduced by -$5,273. But outpatient costs are increased by $10,981 and supported housing costs by $3,180 for a total of $14,161. So despite the reduction in emergency room costs, there is a net increase in costs with housing of $8,888 and without housing of $5,708. To make these service enhancements for the estimated 141,420 persons with SMI in San Diego County Community Health Improvement Partners (2010) might require increasing the mental health budget by $1,256,940,960 with housing costs and $807,225,360 without housing costs. Hardly an insignificant amount. And these are just the increases over expenditures prior to the the FSP.
In theory there might be service packages that reduce the utilization and absolute cost of emergency care without reducing disappearance rates or increasing risk of harm to self or others. Examples might be increased use of jails and prisons or increased hospital use (both in terms of frequency of hospitalization and lengths of stay). However, the total system costs of this trans-institutionalization (including corrections costs) might not be less expensive than current mental health system costs and its human consequences in terms of consumer rights would be undesirable. It might also be possible to simply turn people away from emergency rooms; however this would increase risk of harm to self and others. In other words, the proposed North Carolina solution for reducing emergency service utilization might not work unless quality of service is sacrificed. Thus this may be a case in which, as Levin (1977) put it: “poor quality is the solution, not the problem.”
In summary: when it comes to reducing emergency room costs for persons with SMI, there is probably no such thing as a “free lunch.”
Community Health Improvement Partners (2010). Charting the Course 2010: A San Diego Community Health Needs Assessment.
Gilmer, T. P., et al. (2010). “Effects of full-service partnerships on homelessness, use and costs of mental health services, and quality of life among adults with serious mental illness.” Archives of General Psychiatry 67(6): 645-652.
Levin, G. (1977). “Point of view: poor quality is the solution, not the problem.” Health Care Manage Rev 2(3): 69-72.