Rationality and evidence alone don't win major societal arguments, like the need for health care workers to be competent in geriatric care. But they are the tools we need to build a foundation for change. Add an emotionally powerful story, and you have an opportunity. IHI's campaign for 100,000 Lives (the precursor to its 5 Million Lives Campaign began with some evidence about the deaths caused by quality and safety problems in hospital care (infections, surgical errors, medication errors) and some evidence about strategies to prevent those deaths ("bundles" of interventions including hand washing, protocols for inserting ports, etc.). They then coupled this evidence with an emotionally powerful idea—saving 100,000 lives.

With the publication of the paper “Relationship Between Quality of Care of Hospitalized Vulnerable Elders and Postdischarge Mortality,” published in the September issue of the Journal of the American Geriatrics Society, we now have some of the evidence we need to consider a campaign of our own. Author Vineet Arora, MD, a Hartford Foundation-supported 2006 Health Outcomes Research Scholar, and her colleagues report on an amazing project linking quality of care of hospitalized patients to likelihood of death in the year following discharge. Starting with the ACOVE criteria (Assessing Care of Vulnerable Elders) as modified to study quality of care in hospitalized patients, Dr. Arora looked at percentage of indicated care delivered to vulnerable elders at the University of Chicago hospital through chart review and related it to death as recorded in the national death index.

As usual, the percentage of indicated care delivered according to the patient’s charts was not 100%, but rather averaged 59.3%. This means that patients received only a little better than half of recommended care. The ACOVE hospital indicators are fascinating in themselves. They range from 6 "universals" that apply to all hospitalized elders to things that thankfully applied only to a few patients. Universals include assessment of cognitive status with 24 hours (4% passed), assessment of functional status for new patients (42%), screening for chronic pain (55%), and initiation of discharge planning within 48 hours (86%). Of the condition-specific indicators, five pertain to pressure ulcers (bedsores). The balance have to do with screening for depression among patients with dementia, working up delirium, careful use of restraints, urinary catheters, and -a near universal - recommending physical activity. Given that chart review was the data source, the study could not consider factors we already have reason to believe are important: clear understanding of discharge instructions by patients and families, communication from hospital to post- acute providers, medication reconciliation to prevent medication errors, follow-up, and coaching.

So what did Dr. Arora find? For the 50% of patients 65+ who were assessed as vulnerable, a 10% increase in care quality predicted a 7% reduction in death at the end of one year. Given that 26.7% of the patients died in the year following the hospitalization, this would mean that getting high-quality care has an impact on a very real risk. I know which group I would rather be in, and I suspect that if patients knew, they would have a preference too.

Now, watch as I extrapolate wildly from one study and one hospital to the whole nation and population of older adults and I begin to get really excited. (Although the ACOVE team did actually also demonstrate this type of result in a smaller project reported back in 2005: "Quality of Care Is Associated with Survival in Vulnerable Older Patients" by Higashi T, Shekelle PG, Adams JL, Kamberg CJ, Roth CP, Solomon DH, Reuben DB, Chiang L, MacLean CH, Chang JT, Young RT, Saliba DM, Wenger NS. Annals of Internal Medicine 2005; 143(4):274-281.) By my back-of-the-envelope calculations, if we could increase the quality of care from the mean score at the University of Chicago hospital to 90% of indicated care, we could save 100,000 lives by treating about 7,000,000 vulnerable elders. Or, to rephrase, we would need to treat 70 people at this higher standard of care to save one life. Given rates of hospitalization, I figure that this many vulnerable elders are hospitalized at least once in just two years time. I don't think IHI had any more evidence than this when it started its 100,000 Lives campaign. If they can do it, why can't we?