2012 Potpourri XXVII

A study reported in the British Medical Journal of Quality and Safety looks at medication administration errors.  Based on a retrospective chart review, in over 14,000 medication administrations, there were 1271 errors, out of which 133 had the potential to cause serious or life-threatening harm.  Out of these, ten actually did cause harm, three serious harm and one a life-threatening injury.  Half of the problems were related to anti-hypertension drugs, largely due to dosage or monitoring errors.  The study suggests that given that millions of doses of drugs are administered in hospitals every year; there are a large number of adverse events that could be avoided by eliminating medication errors.  (BMJ Study)

The Journal of the American Medical Association published a study of physician patient-sharing networks, which has relevance for the design and implementation of accountable care organizations and bundled or episode payment systems.  The study examined relationships among doctors treating Medicare beneficiaries.  There was a significant amount of variation in characteristics across hospital referral regions.  The centrality of primary care physicians varied by a factor of 5.  Physicians were most likely to patient-share if they practiced at the same hospital and had offices in closer geographic proximity.  Connected doctors also had similar patient panel demographic and illness burden characteristics.   (JAMA Article)

The Congressional Budget Office once again looks at what is going to happen to the federal budget when the Medicare physician payment issue is finally resolved.  You may recall that projected health spending numbers are significantly fudged because the CBO has to keep pretending that a massive cut is coming.  The CBO looked at a cliff option, where the initiation of the full cut may be delayed but it happens in one or two years; a clawback option whereby the cut is implemented gradually by reductions over ten years or more and the most realistic, an option that forgives the cumulative sustainable growth rate deficit and replaces that formula with something else.  The first option has little effect; the second results in no long-term additional costs, although spending and deficits will be higher until the clawback is completed, and what is really going to happen is going is going to add hundreds of billions of dollars to spending and the deficit.   (CBO Report)

A study presented in Cancer examined the costs and benefits from widespread gene expression testing, which is a primary feature of personalized medicine.  The study looked at use of tumor profiling to guide treatment in breast cancer, using two commonly available profiling tests that help determine which women should get certain therapies.  The costs included all testing and treatment costs, including for adverse events.  Both tests were related to over $20,000 in spending, but added over 7 quality-adjusted life years for a patient.  One test was notably less expensive than the other in terms of cost per QALY, but both seem well-justified in terms of the cost-benefit ratio.  (Cancer Article)

Pay for performance and value-based purchasing rely on accurate measurements of quality.  Even basic measures like mortality can give very different results when measured with different formulas.  Research in the Journal of the American Medical Association looks at different 30-day mortality models for ischemic stroke when evaluating hospital performance.  One model adjusted for stroke severity and the other did not.  Examining data across 782 hospitals, the authors found that the model which included stroke severity was far more accurate and that among the top 20% or bottom 20% of hospitals in an unadjusted model, 27% were reclassified under the adjusted model, with 57% of hospitals having worse than expected scores in the unadjusted model being changed to as expected under the adjusted one.  Shows how dangerous the whole quality measurement business is.   (JAMA Article)