Quality

Recently, the San Diego Union Tribune reported that the Sharp Grossmont Hospital in eastern San Diego county was cited for a number of preventable deaths. Reporter Cherl Clark found numerous problems, which included:

“staff members restraining a highly medicated, 25-year-old man with schizophrenia in such a way that he was allowed to suffocate. In addition, hospital workers caused the death of an 83-year-old woman who had undergone a hysterectomy by injecting a dangerous anti-narcotic into her bloodstream. Other problems included nurses who did not know or use proper CPR, an unsanitary operating-room mattress held together by tape and glue, unsafe storage and handling of food and kitchen equipment, and use of critical medications such as heparin that had expired up to a year earlier.”

CMS is threatening Sharp Grossmont that it could lose all federal money (i.e., Medicare and Medicaid). Since 50% of Sharp Grossmont’s business comes from Medicare and Medicaid patient, this would be a disaster for the hospital. What should we do with underpreforming hospitals?

In regular markets, when a product has lower quality, people stop buying that product and switch to another one. For instance, if GM stops making high quality cars, people switch to Toyota. The market compels companies to offer a desirable bundle of quality and price or else they will lose business. Because markets are so effective at maintaining high quality, withholding Medicare and Medicaid payments from Sharp Grossmont makes sense, right?

Maybe not in this case. First of all, some hospitals may not be in a competitive market. While urban hospitals must compete with other, nearby hospitals, Sharp Grossmont is supposed to provide medical care for the entire Grossmont Healthcare District, which covers 750 square miles in San Diego’s more suburban and rural East County. This area has more than 652,000 residents and Sharp Grossmont has the busiest emergency department in San Diego County. By reducing funding, individuals who have emergencies will receive even worse care than before.

This is similar to the no child life behind program. Low preforming schools lose money. But these are exactly the schools that need more money to survive. If students had the freedom to switch schools, then penalizing a failing school would make perfect sense since the students could opt for higher quality schools. The failure of low quality schools would not be a problem if students had other schools available for them to choose to attend. If individuals do not have any choice of which school they attend, however, withholding funding from schools or hospitals can make low quality schools worse.

Will Sharp Grossmont be decertified? The CMS threat to withhold funding is likely just a bluff.

“Among the 450 hospitals in [CMS certification officer Steven] Chickering’s jurisdiction of Hawaii, California, Nevada and Arizona, 10 to 12 a year have as many major lapses, he said. Ninety-nine percent of those facilities resolve their crises and keep their federal payments, Chickering said.”

The federal government knows that removing funding from the only emergency room in a 750 square mile area is not politically feasible. Although individuals may not have much choice of a hospital in an emergency situation, in non-emergency situations patients can decide to drive longer distances to visit physicians at more competent hospitals. If CMS payments are proportional to patient volume, then Sharp Grossmont may take a financial hit due to this lower patient volume without having the hospital decertified.

Decertification is likely not the answer, but having such serious quality lapses reflects poorly on the state of health care in San Diego, and in the U.S. in general.

• Additional information on Sharp Grossmont’s predicament can be found at KPBS’s the Editors Roundtable.

The Washington Post reports that U.S. Health Care [is] Still Ill. This conclusion comes from a report from the Commonwealth Fund titled Results from the National Scorecard on U.S. Health System Performance, 2008. Even though the U.S. still spends more money on medical care than any other nation, performance on the Scorecard has not improved between 2006 and 2008.

Today, let’s look into some of these measures in detail.

• Preventable Deaths/100,000 Population. This is calculated as the annual ratio of people below age 75 or below who die from diseases such as heart disease, stroke, bacterial infection, diabetes. These figures are age-adjusted so that if one country has a lot of old people, this does not count against them. This is a very good aggregate metric of how well the health care system is doing and it turns out that the U.S. is in last place.
• Percent of adults (18+) who received all recommended screening and preventive care. Here we see pretty much no change between 2002 and 2005 (49% to 50%). However, using a metric such as all preventive care hides some improvements along the margin. For instance, let us assume that there are 5 key vaccines. If and half the population get all the vaccines and half of the population did not, then we would have a 50% score on this metric. On the other hand, if we had a huge improvement where the people who were not getting vaccinated now got 3 of the 5 vaccines, we would still be at the same 50% mark.
• Quality of Care. We note that in most of the measurable quality metrics, performance increased. The number of diabetics with HbA1C<9% increased; the percent of people who received the proper care for heart attacks, heart failure and pneumonia increased as well. We see that more and more patients now receive written instructions after they are discharged from a hospital. Much of this improved quality may be coming from pay-for-performance interventions. As I have mentioned in earlier blog posts, P4P may improve the quality of care on measured dimensions, while reducing the quality of care in unmeasured areas.
• Nursing Home hospitalizations. The percent of nursing home patients who were hospitalized in a year increased. This may be due to worse care, or an older–and thus generally sicker–population of adults being the ones who enter into nursing homes.
• Off-Hours care: Although there was some improvement from 2005, Americans in 2007 were the least likely to be able to receive non-emergency care on nights, weekends, and holidays.
• Access: The percent of people who are uninsured has risen greatly between 2000 and 2006. Further, 41% of adults have an outstanding medical debt of bill problem.
• Coordination of care. When Americans have test done, they are the least likely to have these results available at the time of the next appointment. Among countries in the Commonwealth Report, only Canadians primary care physicians (PCPs) are less likely to use electronic medical records than American PCPs.

Can these report cards help improve care? Yes an no. Of course, pointing out short-comings in the American system is the first step that is needed in order to improve care. The report, however, does not really explain why these short falls are occurring or how to fix them. Are adult preventive care levels low because physicians are not doing their job, or are patients avoiding needed checkups? The answer to this question will determine whether a physician-focused or public health-focused approach would work best. Similarly, we see that nursing home hospitalization are increasing. Why is this? Is this because of worse nursing home care or are the patients who enter into nursing homes sicker on average in more recent years? The the former is the culprit, what specific problems are leading to more hospitalizations and how can we fix them.

I applaud the Commonwealth Fund for collecting this data. As they wisely state, “what receives attention gets improved.” However, more detail studies are needed if we really want to improve the quality, access and efficiency of health care in the 21st century.

Simon Caulkin, management editor of The Guardian, has a great article titled “The rule is simple: be careful what you measure.”  The article discusses the fact that measuring performance leads to better performance on the dimensions measured, but can often lead to significantly worse performance on the unmeasured dimensions.  For instance,

What happens when bad measures drive out good is strikingly described in an article in the current Economic Journal. Investigating the effects of competition in the NHS, Carol Propper and her colleagues made an extraordinary discovery. Under competition, hospitals improved their patient waiting times. At the same time, the death-rate following emergency heart-attack admissions substantially increased. Why? As targets, waiting times were and are measured (and what gets measured gets managed, right?). Emergency heart-attack deaths were not tracked and therefore not managed. Even though no one would argue that the trade-off - shorter waiting times but more deaths - was anything but a travesty of NHS purpose, that’s what the choice of measure produced.

Hat tip to DB’s Medical Rants for this one.

The Retired Doc’s Thoughts has an interesting post as well.

“When regulators or policy makers succeed in improving the quality of care provided by some doctors, do patients even notice and/or care?” This is the question which Kenneth L. Leonard attempts to answer. Davies and Ware (1988) do state that patient satisfaction is correlated with average quality levels. One problem with most studies is that quality improvements take place over a long period of time. It is possible that as quality improves, patient expectations also increase and thus overall reported patient satisfaction may be unaffected.

How does a researcher get an exogenous change in quality? Leonard uses the Hawthorne Effect. The Hawthorne effect states that quality will improve whenever there is a change in environment. In this study, doctors are observed by researchers. The study finds that objective measures of quality improve immediately after researchers arrive, but the quality improvement slowly decays over time until after about 10 observations, the doctor returns to their original quality level.

So, do patients actually notice this quality improvement? Leonard find that:

“Patient do, in fact, recognize and value quality care. A 1% increase in protocol adherence (from an average adherence of about 53%) is associated with about a 0.40% increase in the probability that a patient will declare the consultation to have been “very good” (from an average level of about 12%).”

• KL Leonard (2008) “Is patient satisfaction sensitive to changes in the quality of care? An exploitation of the Hawthorne effect” Journal of Health Economics, 27(2): 444-459.
• A. Davies and J.J. Ware (1988). “Involving consumers in quality of care assessment, Health Affairs” 7, pp. 33–48.