Beginning October 2008, the Centers for Medicare and Medicaid Services (CMS) will no longer reimburse hospitals for costs related to eight common complications, a list compiled with little physician validation and input. With such complications as pressure ulcers making the list, the new policy threatens to put an extraordinary burden on the ED, which may be required to document even the slightest pre-existing co-morbidities.
The eight complications included in the new CMS policy are the following: post-operative retained foreign body, hospital acquired urinary tract infection, central line associated bloodstream infections, incompatible blood products, air embolism, patient falls, post cardiac surgery mediastinal infection, and pressure ulcers. If any of these conditions are identified on discharge that were not present-at-admission, CMS will not reimburse at the higher rate used in years past. For example, a myocardial infarction patient with a UTI developing during their hospitalization will have the same reimbursement rate as an MI patient without this complication despite the extra expense entailed in caring for this complication.
Impact on the hospital
If your hospital receives Medicare or Medicaid funding, this will affect your institution. Early estimates project a $21 million/year immediate reduction in payments for Medicare patients. Beyond the obvious financial ramifications to institutions, these measures may ultimately impact population perceptions when the preventable error rates are made public by CMS. Undoubtedly, below average ratings for individual hospitals will be equated to lower quality care by patients, newspapers, and watchdog groups.
Direct Impact in the Emergency Department
Since these measures are only deemed preventable if they occur during the hospitalization, the onus will be upon the early health care teams to document these conditions from the front door. If a long-standing decubitus ulcer is first documented on hospital day 3, CMS will probably not reimburse for care related to the pressure sore. EM nursing and physician staff will likely be tasked with routinely screening admitted patients for pressure sores, fall risk, and UTIs to enhance the prompt identification of these problems before hospitalization ensues. One unintended consequence of CMS’s preventable errors initiatives may significantly impact EM: high risk populations (obese, elderly, and chronically ill) may encounter increasing difficulty obtaining medical care since hospitals will be disinclined to risk failed reimbursement efforts. Failure to obtain definitive care will drive more of these patients towards the fraying EM safety net at later stages in their disease process, thus increasing already dangerous levels of overcrowding.
Why has CMS decided to withhold reimbursement of known, often unpreventable complications?
Since publication of the IOM’s “To Err Is Human”, most pay-for-performance initiatives have focused upon proactive uptake of evidence-based interventions by financially rewarding compliant providers. Resulting endeavors to promote patient safety have included public education through media reports, pressure from regulatory bodies, and public reporting. While success stories have been reported, many feel that enhancements to patient safety initiatives have been too slow and accomplished with substantial resistance. Turning to pay as a potential motivator, three general methods might be applied: pay more for safer care, promote safe care competition to attract a better payer mix, or pay less for unsafe care. The preventable errors proposal is using the latter approach and originated from the Deficit Reduction Act of 2005. At that time CMS was tasked with identifying two hospital-acquired conditions, but in consultation with the Center for Disease Control and Prevention (CDC) selected eight conditions for their first round of reforms as a “stick” approach to garner the attention of hospitals and health care professionals.
What qualities best define a preventable measure?
To improve quality of care, candidate complications must be important, measurable, and preventable. Broadly defined, importance is first suggested by high volume and high cost burdens to society. The current measures are certainly important, but depending upon one’s practice perspective, most emergency physicians would probably not agree on whether they are the most important. In reducing adverse patient outcomes, explicit, transparent, reproducible methods of quantifying importance need to be developed. Unfortunately, accurately measuring the institutional and regional prevalence of a diagnosis remains a challenge for almost any condition. Identifying problems will depend in part on how hard one looks. One hospital’s attempts to identify every case which they are then mandated to report, may be represented as deficiencies by the public when comparing numbers with competing institutions who self-report lower rates simply because they do not choose to look as carefully. Even if near-perfect diagnoses can be confidently attained, attributing a cause-effect relationship between provider error and undesired outcome will be a tenuous proposition for the sustainable future. Finally, prevention must be generally attainable through realistic practice change. If a measure cannot be efficiently prevented, who benefits by early detection? Even if one condition can be prevented, how many will suffer the iatrogenic complications of proactive treatment so that one might benefit?
Defining a “preventable error”
Beginning in January 2008, all Medicare discharge claims had to include a present-on-admission indicator for all secondary conditions. If your hospital is billing Medicare for a diagnosis, your billers have to clearly distinguish conditions existing prior to admission. Other insurers are expected to follow the CMS lead. However, the Leapfrog Group envisions error recognition as only the first step. In addition, for “never” events like leaving a sponge in a patient following surgery (the National Quality Forum lists 28 “never events”), they advocate a minimum institutional response to include the following:
1) An official apology to the patient and family
2) Reporting the event to at least one formal body
3) Conduct a root-cause analysis for each event
4) Waiving all costs directly related to an event while refraining from seeking third party payments.
For example, consider the diagnosis of UTI. Indwelling urinary catheters account for 80% of nosocomial UTI’s and 40% of all nosocomial infections costing the US healthcare system $400 million annually. A patient presenting with congestive heart failure often requires diuresis. A urinal or bedpan can be used to accurately measure output, although many patients find these devices offensive and unacceptable. Instead, patients request bathroom privileges and some fall in traveling from bed to bathroom, either in the ED or once admitted to the hospital. A fraction of these falls will be injurious and complicate the hospital course. Is a clinician to avoid short-term urinary catheterization to minimize iatrogenic UTI at the risk of inaccurate urine output measures? Alternatively, providers could allow patients to ambulate to the bathroom, but then those suffering a complicated fall will be similarly identified as a preventable error. Should every patient receive pre-hospitalization testing for a pre-existing UTI? Since the criterion standard is urine culture, should every patient receive a urine culture upon admission to fully document infections present at the time of admission? Doing so would certainly increase costs and diminish quality for the many false positives. In fact, many diagnostic tests we use lack perfect sensitivity and specificity and will therefore yield large numbers of false-negatives and false-positives, particularly as more asymptomatic individuals are screened. In addition, how many false-positive patients will suffer adverse drug reactions from unnecessary antibiotic exposures?
In support of the CMS approach, the first CDC guideline for the prevention of catheter-related UTIs was published in 1981 and has remained widely ignored. The new CMS policy linking preventable errors to hospital reimbursement will undoubtedly serve as a paradigm shift when physicians, nurses, and ancillary providers are faced with the fiscal consequences of inaction. Whether the intended consequence (reduced catheter-associated UTI rates) outweigh the inevitable unintended consequences (unnecessary antibiotic exposures, distrust in future Quality Improvement measures) remains to be seen.
The CMS’ “reasonably preventable errors” list has been implemented with very little validation, physician input, or health care provider education. Without waiting to assess the unintended consequences of these proactive patient safety measures, CMS has now proposed an expanded preventable errors list in 2009 to include ventilator-associated pneumonia, S. aureus septicemia, venous thromboembolism, delirium, extreme blood sugar derangements, iatrogenic pneumothorax, and C. difficile colitis. Reducing reimbursement rates will undoubtedly attract the attention of hospital administrators who will in turn task providers with the development of institutional approaches to reduce such losses. As the front door to the hospital, EM will be charged with detailed documentation of all pre-existing co-morbidities including UTI’s and bed sores for all admitted patients. If CMS subsequently expands the list of preventable errors, EM will be further burdened with such screening. In addition, advocates for preventable errors payment withholdings are now calling for the physicians’ portion of fees to be the next step in this process. By monitoring and reporting the real-world impact of these measures on the emergency care of your patients, you can help attain the goal that all parties share in a chaotic environment: optimal patient outcomes.
Case Study: Diagnosing UTI
Indwelling urinary catheters account for 80% of nosocomial UTI’s and 40% of all nosocomial infections costing the US healthcare system $400 million annually. A patient presenting with congestive heart failure often requires diuresis. A urinal or bedpan can be used to accurately measure output, but many patients find these devices offensive and unacceptable. Instead, patients request bathroom privileges and some fall while traveling from bed to bathroom, either in the ED or once admitted to the hospital. A fraction of these falls will be injurious and complicate the hospital course. Is a clinician to avoid short-term urinary catheterization to minimize iatrogenic UTI at the risk of inaccurate urine output measures?