Are We Killing People With 30-Day Readmission Targets?

Ever since the Center for Medicare and Medicaid Services announced their intention to penalize hospitals for early readmissions, folks have been worrying about the obvious consequences: would a focus on avoidance place patients at risk? Would patients best served in the hospital be pushed into other settings for suboptimal care?

That is the argument made in this short piece in the Journal of the American College of Cardiology. They look backwards at the last two decades of heart failure readmissions and short-term mortality, and take issue with the fundamental underlying premise of the quality measure, the inequities associated with the measure, and potential unintended harms. Their most illustrative example: when patients die outside the hospital within 30-days, paradoxically, they contribute to apparent improved performance in healthcare quality, as measured by 30-day readmission.

They back up their point by using the aggregate data analyzing readmissions between 2008 and 2014, published previously in JAMA, and focusing primarily on the heart failure component. In the original JAMA analysis, the evaluation paired individual hospital monthly readmission and risk-adjusted mortality, and were unable to identify an increased risk of death relating to reductions in 30-day readmissions. These authors say: too much tree, not enough forest. In the decade prior to announcements of 30-day readmission penalties, 30-day heart failure mortality had dropped 16.2%, but over the analysis period, 30-day heart failure mortality was back on the rise. In 2008 the 30-day mortality was 7.9% and by 2014 it was up to 9.2%, a 16.5% increase, and an even larger increase relative to the pre-study trend with decreasing mortality.

These are obviously two very different ways of looking at the same data, but the implication is fair: those charged with developing a quality measure should be able to conclusively demonstrate its effectiveness and safety. If any method of analysis raises concerns regarding the accepted balance of value and harm, the measure should be placed on a probationary status while rigorous re-evaluation proceeds.

“The Hospital Readmission Reduction Program Is Associated With Fewer Readmissions, More Deaths”

Is 3.3% Really the Avoidable Number?

Much has been made about “avoidable” Emergency Department visits – something, as physicians, we are anecdotally all-too-familiar – but the exact scope of the problem has been difficult to quantify. If it were easy, after all, the range of estimates for avoidable visits would not encompass the spread between 4.8% and 90%.

This is another shot at defining avoidable, this time using a very restrictive definition. An avoidable ED visit occurs when a patient is discharged home, and:

  • No tests are performed
  • No procedures are performed
  • No medications are administered or prescribed

These authors utilize the National Hospital Ambulatory Medical Care Survey (NHAMCS) for years 2005–2011, and find an estimated 3.3% of ED visits met those criteria. The most common discharge diagnoses included those related to alcohol use, back pain, throat pain, upper respiratory symptoms, dental issues, and mental health issues.

But, the true net result of this analysis is, unfortunately, it simply grows our estimate for avoidable ED visits to an even-more-ridiculous range between 3.3% and 90%. This definition lends itself far more to convenience for data analysis than face validity as a surrogate for avoidable. Clearly, a lack of testing or medication administration does not indicate a life- or limb-threatening condition has not been ruled out by expert clinical examination and reasoning. Conversely, ordering a test or administering a medication does not indicate the level of service of an ED was necessary, or the ideal venue for care delivery.

Given the limitations of the data set, this is a reasonable approach to add to the discussion of the types of potentially avoidable ED visits. However, I expect to see this 3.3% number cited frequently in isolation without acknowledging the underlying definitions or methods for derivation.  A better title for this paper?  How about: “ED visits resulting in neither treatment nor testing: a descriptive analysis” – the word “avoidable” should be omitted.

“Avoidable emergency department visits: a starting point”

The Futility of NSAIDs for Back Pain?

This article filled with reproach for non-steroidal anti-inflammatories was highlighted in a New England Journal of Medicine Journal Watch and on Twitter – a wistful treatise remarking on the general ineffectiveness of pharmacologic analgesics. “Nothing works!” accompanied by a general gnashing of teeth and writhing on invisible flames.

But – does this meta-analysis actually reach such a conclusion? Examine the first few words in their conclusion:

NSAIDs are effective for spinal pain …

Off to a good start! But, the catch:

… but the magnitude of the difference in outcomes between the intervention and placebo groups is not clinically important.

These authors pool the results of 35 randomized, placebo-controlled trials for “spinal pain”, which is to say undifferentiated pain relating anatomically to any part of the spine. These trials comprised 6,065 participants – or, if you do the math, an average of 173 patients per trial, nearly all of them performed over a decade ago. The pooled effects of these trials all favored NSAIDs – but, as the authors mention, the absolute magnitude of effect on pain scales was a the edge of their threshold for clinical significance. The authors defined a difference of 10 points on a 100-point scale as clinically important, but most of their pooled results landed between -7 and -16, favoring NSAIDs over placebo. With these small samples, generally moderate GRADE quality, and moderate to high heterogeneity between the pooled results, there is a lot of fuzziness around their ultimate conclusion.

These authors do many, other, exploratory analyses, and it is reasonable to suggest the limitations inherent to each render any conclusions unreliable. Adverse events, as reported, were similar between groups – excepting for increased gastrointestinal adverse events, most of which were non-serious. The authors report this difference as a relative risk of 2.5 for GI side effects in their comparison, but the absolute differences are on the order of an excess of 1 in 100.

This is probably much ado about nothing. Their perspective is not inaccurate, per se, but these trials do find a consistent benefit to NSAIDs. The value judgment here on clinical effectiveness probably misses the mark, particularly considering these are inexpensive, readily available, with few adverse effects in short-term use. I would probably argue it is easier to defend a position they still have utility in multi-modal pain control regimens, rather than to conclude they be consigned to the rubbish bin.

“Non-steroidal anti-inflammatory drugs for spinal pain: a systematic review and meta-analysis”


Blood Cultures Save Lives and Other Pearls of Wisdom

It’s been sixteen years since the introduction of Early Goal-Directed Therapy in the Emergency Department. For the past decade and a half, our lives have been turned upside-down by quality measures tied to the elements of this bundle. Remember when every patient with sepsis was mandated to receive a central line? How great were the costs – in real, in time, and in actual harms from these well-intentioned yet erroneous directives based off a single trial?

Regardless, thanks to the various follow-ups testing strict protocolization against the spectrum of timely recognition and aggressive intervention, we’ve come a long way. However, there are still mandates incorporating the vestiges of such elements of care –such as those introduced by the New York State Department of Health. Patients diagnosed with severe sepsis or septic shock are required to complete protocols consisting of 3-hour and 6-hour bundles including blood cultures, antibiotics, and intravenous fluids, among others.

This article, from the New England Journal, looks retrospectively at the mortality rates associated with completion of these various elements. Stratified by time-to-completion following initiation of the 3-hour bundle within 6 hours of arrival to the Emergency Department, these authors looked at the mortality associations of the bundle elements.

Winners: obtaining blood cultures, administering antibiotics, and measuring serum lactate
Losers: time to completion of a bolus of intravenous fluids

Of course, since blood cultures are obtained prior to antibiotic administration, these outcomes are co-linear – and they don’t actually save lives, as facetiously suggested in the post heading. But, antibiotic administration was associated with a fraction of a percent of increased mortality per hour delay over the first 12 hours after initiation of the bundle. Intravenous fluid administration, however, showed no apparent association with mortality.

These data are fraught with issues, of course, relating to their retrospective nature and the limitations of the underlying data collection. Their adjusted model accounts for a handful of features, but there are still potential confounders influencing mortality of those who received their bundle completion within 3 hours as compared to those who did not.  The differences in mortality, while a hard and important endpoint, are quite small.  Earlier is probably better, but the individual magnitude of benefit will be unevenly distributed around the average benefit, and while a delay of several hours might matter, minutes probably do not.  The authors are appropriately reserved with their conclusions, however, only stating these observational data support associations between mortality and antibiotic administration, and do not extend to any causal inferences.

The lack of an association between intravenous fluids and mortality, however, raises significant questions requiring further prospective investigation. Could it be, after these years wandering in the wilderness with such aggressive protocols, the only universally key feature is the initiation of appropriate antibiotics? Do our intravenous fluids, given without regard to individual patient factors, simply harm as many as they help, resulting in no net benefit?

These questions will need to be addressed in randomized controlled trials before the next level of evolution in our approach to sepsis, but the equipoise for such trials may now exist – to complete our journey from Early Goal-Directed to Source Control and Patient-Centered.  The difficulty will be, again, in pushing back against well-meaning but ill-conceived quality measures whose net effect on Emergency Department resource utilization may be harm, with only small benefits to a subset of critically ill patients with sepsis.

“Time to Treatment and Mortality during Mandated Emergency Care for Sepsis”

Just the Cost of Doing Business

Good news, everyone!

In the past two decades, for virtually every specialty, the number of paid medical malpractice claims has decreased. Overall, for all specialties, the rate of payment has been halved, compared with the 1992-1996 timeframe. Neurosurgery, unfortunately, is still the “winner”, followed by plastic surgery, thoracic surgery, and obstetrics. The lowest rates were seen in psychiatry and pediatrics. Emergency medicine sits right in the middle, with 18.8 paid claims per 1,000 physician years.

The bad news, unfortunately, was that the claim amounts – including paid claims greater than $1 million – increased. Emergency Medicine paid claim amounts increased 26.1% to a mean of $314,052 in the most recent time period of analysis, an increase in line with the overall mean for all specialties. The largest jump in payout amounts was essentially a tie between dermatology, gastroenterology, pathology, and urology. Neurosurgery actually had one of the lowest payout increases – probably because they started from such lofty heights, already.

Types of malpractice alleged varied by specialty, with the expected variation between diagnostic error, surgical error, and treatment errors between the diagnostic and surgical specialties. Most (63.6%) of malpractice alleged in emergency medicine fell into alleged diagnostic error, while logically 73.3% of alleged error in plastic surgery fell under surgical error.

These data, from the National Provider Data Bank, only document payments made for written claims and do not include settlements or monies paid out by institutions. Whether these actually represent a friendlier environment for physicians, more aggressive approaches to settling claims, or a shifting of liability to corporate proxy is not clear. Regardless, even if it is a little of all three, the trend is probably moving in the right direction.

“Rates and Characteristics of Paid Malpractice Claims Among US Physicians by Specialty, 1992-2014”

Discharged and Dropped Dead

The Emergency Department is a land of uncertainty. Generally a time-compressed, zero-continuity environment with limited resources, we frequently need to make relatively rapid decisions based on incomplete information. The goal, in general, is to treat and disposition patients in an advantageous fashion to prevent morbidity and mortality, while minimizing the costs and other harms.

The consequence of this confluence of factors leads, unfortunately, to a handful of patients who meet their unfortunate end following discharge. A Kaiser Permanente Emergency Department cohort analysis found 0.05% died within 7 days of discharge, and identified a few interesting risk factors regarding their outcomes. This new article, in the BMJ, describes the outcomes of a Medicare cohort following discharge – and finds both similarities and differences.

One notable difference, and a focus of the authors, is that 0.12% of patients discharged from the Emergency Department died within 7 days. This is a much larger proportion than the Kaiser cohort, however, the Medicare population is obviously a much older cohort with greater comorbidities. Then, they found similarities regarding the risks for death – most prominently, “altered mental status”. The full accounting of clinical features is described in the figure below:

Then, there were some system-level factors as well. Potentially, rural emergency departments and those with low annual volumes contributed in their multivariate model to increased risk of death. This data set is insufficient to draw any specific conclusions regarding these contributing factors, but it raises questions for future research. In general, however, this is interesting – and not terribly surprising data – even if it is hard to identify specific operational interventions based on these broad strokes.

“Early death after discharge from emergency departments: analysis of national US insurance claims data”

Insight Is Insufficient

In this depressing trial, we witness a disheartening truth – physicians won’t necessarily do better, even if they know they’re not doing well.

This study tested a mixed educational and peer comparison intervention on primary care physicians in Switzerland, with an end goal of improving antibiotic stewardship for common ambulatory complaints. The “worst-performing” 2,900 physicians with respect to antibiotic prescribing rates were enrolled and randomized to the study intervention or none. The study intervention consisted of materials regarding appropriate prescribing, along with personalized feedback regarding where their prescribing rate ranked compared to the entire national cohort. The core of their hypothesis involved whether just this passive knowledge regarding their peer performance would exert normalizing influence over their practice.

Unfortunately, despite providing these physicians with this insight, as well as tools for improvement, the net effect of their intervention was effectively zero. There were some observations regarding changes in prescribing rates for certain age groups, and for certain types of antibiotics, but dredging through these secondary outcomes leads to only unreliable conclusions.

This is not particularly surprising data. These sorts of passive feedback mechanisms unhitched from material consequences have never previously been shown to be effective. There are other, more effective mechanisms – focused education, decision-support interventions, and shared decision-making – but, for a fragmented, national health system, this represented a relatively inexpensive model to test.

Try again!

“Personalized Prescription Feedback Using Routinely Collected Data to Reduce Antibiotic Use in Primary Care”

The Downside of Antibiotic Stewardship

There are many advantages to curtailing antibiotic prescribing. Costs are reduced, fewer antibiotic-resistant bacteria are induced, and treatment-associated adverse events are eliminated.

This retrospective, population-based study, however, illuminates the potential drawbacks. Using electronic record review spanning 10 years of general practice encounters, these authors compared infectious complication rates between practices with low and high antibiotic prescribing rates. Spanning 45.5 million person-years of follow-up after office visits for respiratory tract infections, there is both reason for reassurance and reason for further concern.

On the “pro” side, cases of mastoiditis, empyema, bacterial meningitis, intracranial abscess and Lemierre’s syndrome were no different between those who prescribed high rates (>58%) and those with low rates (<44%). However, there is a reasonably clear linear relationship with excess follow-up encounters for both pneumonia and peritonsilar abscess. Incidence rate ratios were 0.70 compared with reference for pneumonia and 0.78 compared with reference for peritonsillar abscess. However, the absolute differences can best be described as “large handful” and “small handful” of extra cases per 100,000 encounters

There are many rough edges and flaws relating to these data, some of which are probably adequately defeated by the massive cohort size. I think it is reasonable to interpret this article as accurately reflecting true harms from antibiotic stewardship. More work should absolutely be pursued in terms of strategies to mitigate these potential downstream complications, but I believe the balance of benefits and harms still falls on the side of continued efforts in stewardship.

“Safety of reduced antibiotic prescribing for self limiting respiratory tract infections in primary care: cohort study using electronic health records”

How Many ED Visits are Truly Inappropriate?

I’ve seen quite a bit of feedback on social media regarding this research letter in JAMA Internal Medicine.

This study evaluated, using National Hospital Ambulatory Medical Care Survey data, the incidence of hospital admission stratified by triage Emergency Severity Index.  They analyzed 59,293 representative visits from the sample and found 7.5% of them, on a weighted basis, were categorized as “non-urgent” – an ESI level 5 or presumed equivalent.  The typical assumption regarding these non-urgent visits is they represent inappropriate Emergency Department utilization.  This study found, however:

“… a nontrivial proportion of ED visits that were deemed nonurgent arrived by ambulance, received diagnostic services, had procedures performed, and were admitted to the hospital, including to critical care units.”

There are always limitations regarding the NHAMCS data, particularly with missing and imputed data.  Based on this, I tend to feel these data lack face validity.  The weighted incidence of admission for non-urgent patients was 4.4% compared with 12.8% of urgent visits, while 0.7% of non-urgent visits were to critical care units compared with 1.3% of urgent visits.  I certainly do not see similar relative proportions of admission, and then to critical care, for level 5 patients in my multiple practice environments.

Regardless, the general implication made by these authors is probably reasonable, refuting usage of ESI triage level 5 to accurately represent inappropriate Emergency Department visits.  However, left equally unstated, is an acknowledgement that ESI also fails to accurately categorize urgent visits – which ties to the rhetoric of trying to conflate “non-urgent” as “inappropriate and “urgent” as “appropriate”.

ESI, as currently implemented, will not be a reliable tool for directing patients to other sources of care – but, with some fuzziness, probably still gives a reasonable estimate of the overall burden of inappropriate ED visits for some policy applications.

“Urgent Care Needs Among Nonurgent Visits to the Emergency Department”

Too Many Tests! Or, So We Believe ….

Yes, Virginia, we order too many tests.  And, we know it – as evidenced by such conferences on overdiagnosis and costs of care.  And, even more relevant than such academic exercises, as this study indicates, even the general clinician seems to have a fair bit of self-awareness.

In this survey consisting of 435 respondents, 85% of emergency physicians believed excessive testing occurred in their Emergency Department.  Most frequently, such testing was motived by fear of missing even rare diagnoses, but defensive medicine and malpractice came a close second.  Patient expectations, local practice patterns, and time saving were also substantially cited as motivators for ordering.  Thankfully, administrative and personal motivations to increase reimbursement were rarely reported as reasons.

Despite the protestations of some policy-makers, the clinicians surveyed believed the most helpful change to the system would be malpractice reform.  Interestingly, the next ranked helpful interventions included educating patients and increasing shared decision-making.  While the first item may be logistically (or politically) unachievable, there are no obstacles to integrating improved communication behaviors into routine practice.  It does, however, show a need for increased availability of tools for clinicians to use at the point of care.

There are flaws in these sorts of perception-based surveys with regard to the accuracy of such anecdotal self-assessment.  Physician assessment of their own practice and that of others can certainly be questioned.  It must be admitted, however, a more intensive just-in-time surveying method would likely impact the variables measured.

There are also some highly entertaining outliers in Figure 2, of course, the perception of self vs. colleague ordering.  There is a handful of physicians who believe they, themselves, order over 80% of their CTs and MRIs unnecessarily – but that no one else in their group does.  Likewise, there is a handful with just the opposite perception – that their colleagues over-order, while they, themselves rarely do.  I wonder if they work in the same department?

Regardless, first step is admitting you have a problem.  We have many steps yet to go.

“Emergency Physician Perceptions of Medically Unnecessary Advanced Diagnostic Imaging”