Reducing readmissions

Reducing readmissions

March 2014
SHARE

Published in the March 2014 issue of Today’s Hospitalist

GROUCHO MARX, A MASTER OF REPARTEE, ONCE SAID, “A hospital bed is a parked taxi with the meter running.”

Until the Affordable Care Act established the hospital readmissions reduction program, the meter would just keep climbing “even when discharged patients bounced right back. With nearly one in five elderly patients readmitted within 30 days, Medicare’s readmission price tag has been pegged at $26 billion a year.

Regardless of whether readmissions are preventable or what the link may be between readmissions and quality, the high cost of readmissions is undeniable. As a result, financial penalties are now a reality.

A rush to judgment
I attended a hospital-arranged meeting among inpatient and outpatient clinicians to discuss readmissions. One primary care physician claimed to know the truth: Readmissions were the result of “poor discharge summaries” by the hospitalists.

It made perfect sense to everyone in the room “except to the badly outnumbered hospitalists. When asked about the hospital’s readmission rates, administrators repeated national statistics along with their mantra for the hospitalists to do better.

Only later did we discover that the hospital’s readmission rate had continuously improved over the previous three years and was 6.5 percentage points below the national benchmark, putting the hospital in the top 16th percentile.

“The supposition is prevalent the world over that there would be no problems in production or in service if only our production workers would do their jobs in the way that they were taught. Pleasant dreams. The workers are handicapped by the system, and the system belongs to the management.” Many administrators have yet to learn this profound lesson from quality guru W. Edwards Deming, which he wrote more than 30 years ago.

If hospital management owns the system that produces readmissions, what should management’s focus be? Understanding the extent and nature of readmissions would be a good start.

One size doesn’t fit all
Now that readmissions have become a serious pocketbook issue, consultants are falling over each other to sell a bag of generic goods. A new study every month seems to identify additional readmission risk factors.

Readmissions rarely have a single cause or effect. That is why blindly copying generic best practices or research studies “without properly understanding the logic, context, relevance or implications of those practices “is like writing the same prescription for the next patient because it worked so well for your last one. The profiles of readmitted patients at a large, urban hospital may be quite different from those at a smaller, rural hospital. And an inner city hospital may have its own unique challenges compared to one in the suburbs.

Like politics, all readmissions may be local.

Evidence-based management
Studies have reported a confounding mix of factors behind readmissions, including everything from medical complexity to patients’ socioeconomic status, sodium and hemoglobin levels at discharge, and whether the admission was elective or not.

Here is a non-exhaustive starter list of factors to consider to better target readmission-prevention strategies: age, gender, race and ethnicity, illness severity, length of stay, comorbidities, leaving AMA, Medicaid/dual eligibility, delayed follow-up, ED use, and language barriers.

And finally: psychosocial factors. Certain factors “living alone, tobacco use and/or alcohol use, telehealth enrollment, the use of visiting nurse services, and depression or anxiety “significantly increase predicted readmission risk.

Gaining actionable insights
Where to start? In situations affected by complex factors, systematic statistical analysis using business analytics software can help isolate useful insights. Here are six basic steps to follow:

  • Identify specific, independent predictor variables for each patient discharge, using the list above as a guide. Make the list of variables as comprehensive as possible.
  • Collect (or extract) all independent variables for each patient discharge over a long enough period of time, along with a target dependent variable, such as “readmit status” defined as yes or no.
  • Use business analytics software (such as SPSS Decision Trees from IBM) to study the relationship between the dependent variable “in this case, readmit status “with all the independent variables such as age, gender, etc.
  • Generate high-risk readmission profiles by creating a tree diagram that divides the overall data into sub-segments. (Examples of decision trees are online.)
  • Zero in on “hot spots,” which are those segments of your patient population with disproportionate readmission levels, and ask “why?” Engage key stakeholders to drill down and establish customized process-improvement protocols across the entire continuum of care, from admission to pre-discharge and post-discharge follow-up.
  • Flag each high-risk patient who presents and follow the established protocols. Continue to gather data and recalibrate your high-risk readmission profiles and protocols.Show me the system
    When trying to make do without good information or collaborative problem-solving, hospitals resort to overhyped bromides, meaningless goals, a political blame-game or single interventions that may work in studies but don’t really address the problems within a particular hospital or system.

    Here’s an example: An analysis of readmitted patients at one hospital found that nearly one-half of them had no (or a non-standard) primary care provider listed in the hospital’s system. Although an automated hospitalist-to-primary care physician fax notification protocol was in place, that fix was useless in terms of alerting patients’ non-existent primary care physicians or ensuring follow-up appointments within seven days of discharge. Some patients listed a nurse practitioner, for instance, but faxes went out only to doctors.

    A systematic review of interventions has shown that no single solution, implemented alone, is regularly associated with reduced risk of 30-day rehospitalizations. Hospitalists need to urge hospital administrators to set aside feel-good readmission goals and take ownership of their own systems instead. Targeted insights and customized interventions that fit the diagnosis, based on comprehensive data, will be far more effective in reducing readmissions than hit-or-miss fads.

    Abhay Padgaonkar is an award-winning management consultant and an expert in hospital medicine. He is the president of Innovative Solutions Consulting and can be reached via e-mail at abhay@pobox.com. He welcomes the opportunity to help reduce readmissions by using a collaborative, data-driven approach.

NO COMMENTS