Blog Home


Where Was The Leadership? The Questions Raised By Jonathan Welch’s Narrative Matters Essay

May 21st, 2013

Dr. Jonathan Welch’s Narrative Matters essay in the December, 2012 edition of Health Affairs, regarding the cascade of errors and omissions he witnessed in connection with the care provided to his mother, should raise profound questions about how the hospital allowed those failures of care to happen.  Dr. Welch, an emergency medicine physician, watched helplessly as his mother received indifferent care from various nurses and doctors and ultimately died.  Despite having classic signs of evolving sepsis, she was not closely monitored by the nursing staff which ignored alarming signs, was not put on a sepsis treatment protocol by her oncologist, and was not put in an intensive care unit where she could receive more intense monitoring and aggressive treatment from specialists.

While it is tempting to blame the nurse (for not taking vital signs frequently enough and not reacting to abnormal vital signs) and the oncologist (for not following the patient closely enough, not initiating appropriate treatment, and not involving other specialists), Dr. Welch’s story suggests that there were more deeply rooted systemic problems at the hospital that went beyond the shortcomings of the individuals involved in his mother’s care.

As health care attorneys who represent hospitals and physicians, we believe there are some fundamental questions which should be asked by this hospital’s administration, medical staff leadership and governing body to ensure Dr. Welch’s experience is not repeated.  Those questions, which the leaders in all hospitals should consider, include the following:

1.  Does the hospital have political or cultural barriers that would interfere with a patient’s receiving appropriate care, such as being transferred to a higher level of care (e.g. ICU) and receiving treatment from an appropriate specialist (e.g. intensivist)?

As the practice of medicine becomes more specialized, it is not uncommon for there to be “political” barriers to improving patient care.  Internal medicine physicians and pulmonologists who have traditionally followed patients on intensive care units by visiting the patients once or twice a day are often resistant to being replaced by full-time, specialty-trained intensivists who are present in the ICU 24 hours a day, 7 days a week and continuously monitor patients. Internists and family practitioners who have made a significant portion of their income from following patients in the hospital often oppose the hospital hiring hospitalists who are always in the hospital and available to provide continuous care to patients.  It is incumbent on hospital leaders to understand whether such barriers exist and, if so, address them.  The economic interests of a few physicians must not interfere with the delivery of the best possible care.

2. Do nurses feel they are able to discuss concerns about patient care with attending physicians?

Whether it is due to a culture of deference to the physician  or the personality of an individual doctor, nurses frequently are reluctant to call a physician at 2 a.m. regarding a patient’s worsening condition, or to confront a physician who is indifferent to a patient’s needs.  Nurses are the health care providers who are most in touch with the patient’s condition and hospital leaders need to empower them to speak out whenever they feel that a patient’s needs are being neglected. Moreover, if nurses do not get satisfactory responses from a physician, they should know how to go “up the chain of command” and have no reluctance to do so.

3.  Would nurse managers have been aware of and acted on this situation as it developed?

Conspicuously absent from Dr. Welch’s story is any discussion of whether the hospital’s nursing managers were aware of the unfolding events and attempted to intervene.  If the bedside nurse had concerns about the patient’s not being placed on a sepsis protocol or being too sick to be on a regular medical/surgical unit, those concerns should have been discussed with the nursing managers who could assist the nurse in mobilizing the appropriate resources.  However, nurse managers should not only be responsive to issues that are brought to them by the bedside nurses.  They should also have systems in place to flag situations that require attention, even if not recognized by the nurse providing the hands-on care.  A crucial question for hospital leaders is whether the nursing managers knew about the events involving Dr. Welch’s mother as they occurred.  If they did, why didn’t they intervene, and if they didn’t know, why?

4.  Would this kind of event have been identified and investigated to determine the root cause of the failure of care and implement remedial measures?

One of the essential hallmarks of a viable quality assessment and improvement program is a vigorous monitoring and reporting program that picks up on the kinds of events that were observed by Dr. Welch. Members of hospital governing bodies should be assured that when untoward events occur, they are detected and addressed to identify the cause and fix the problems.  Moreover, it is essential that appropriate investigations and actions take place, even absent a complaint from the patient or family members.  If appropriate quality monitoring systems are in place, hospital leaders should never have to learn about a situation such as Dr. Welch encountered from the family or the press.

5.  Would the medical staff be willing to initiate peer review and appropriate action regarding a physician who did not appropriately handle a patient’s care such as in this case?

The hospital’s medical staff must have peer review processes which identify inappropriate care provided by physicians, as well as procedures to address such conduct.  Corrective action could range from simple discussions with a physician to terminating a physician’s medical staff membership and clinical privileges.  But galvanizing medical staffs into taking action against their colleagues is often difficult, and as a result marginal or underperforming physicians continue to practice in hospitals.

6.  Is there an established process for reaching out to patients and families who have experienced adverse outcomes, answering their questions and dealing openly and honestly with mistakes that have been made?

All hospitals are expected to have policies and procedures for dealing with “unanticipated outcomes.”  Those policies generally provide for adverse events to be openly and candidly discussed with patients and family members, and disclosed to patients and families even in situations where the patient or family may not know that something went wrong.  Dr. Welch should not have had to initiate contact with the hospital leadership; however, once he did he should have been engaged in a meaningful dialogue that frankly recognized what went wrong and gave Dr. Welch assurance that at the very least his mother’s death would serve as a learning experience that would avoid such unnecessary deaths in the future.  Not only is accepting responsibility for mistakes the right thing to do, both for the patient and the culture of the hospital, studies have shown that a proactive approach to untoward events often avoids or minimizes malpractice claims.

Dr. Welch’s story highlights a series of systemic failures where the health care system did not perform as it should have on multiple levels.  Hospitals have for years been using a tool know as a “root cause analysis” (affectionately known by some as a “root canalysis” because of the sometimes painful information it produces), which is designed to ferret out the real reasons behind a failure to deliver quality health care.  A root cause analysis in this case might answer many questions about why events unfolded as they did, such as:

  • Did the nursing staff not take vital signs frequently enough, or not recognize the significance of abnormal signs, because they were poorly trained, understaffed or overworked, or for some other reason?
  • Did the nurses not call the oncologist in the middle of the night to suggest to him that a sepsis protocol should be implemented, or that the patient should be transferred to the ICU, because the oncologist was known to berate and humiliate nurses who bothered him at night or dared to challenge his care?
  • Were standard protocols for treating conditions such as sepsis, stroke, and heart attacks not used because they were poorly designed or because physicians were resistant to what they considered (erroneously) to be “cookbook” medicine?
  • Was the involvement of appropriate specialists delayed or avoided because the attending physician was more interested in being able to continue to provide (and bill for) care than he was in the patient’s receiving the best available care?

Hopefully, Dr. Welch’s story will lead to continuous quality improvements not only in the hospital where his mother was treated but in hospitals across the country.  Hopefully, members of hospital governing bodies and administrative and medical staff leaders will read his story and begin asking questions about whether such a sequence of events could occur in their hospital.  Hospital board members should ask themselves, “If I were asked by a friend or colleague whether this could happen in my hospital, would I be able to answer ‘no?’  More importantly, would I be able to answer the follow up question: ‘How do you know?’”

Email This Post Email This Post Print This Post Print This Post

 to the #1 source of health policy research.

No Trackbacks for “Where Was The Leadership? The Questions Raised By Jonathan Welch’s Narrative Matters Essay”

3 Responses to “Where Was The Leadership? The Questions Raised By Jonathan Welch’s Narrative Matters Essay”

  1. Thomas Cox Says:

    Alan: Wouldn’t it begin to make sense to ask why these reports are not being made, how it ties in with hiring cheaper physicians, less qualified nurses, and boosting the responsibilities of unlicensed staff?

    Hospitals are not making these reports for good reasons – unqualified physicians permeate the health care system. Rooting them out, barring them from practice and improving the quality of care would be the appropriate response and yet despite the concentration of thousands of hospitals in the hands of a few, very large health care systems, what seems to make a good deal of sense is not occurring.

    Poor practitioners are more beholden to their employers than well qualified physicians. Same for poor nurses and other health care workers.

    Try to focus on the root causes of the quality problem, not just the most obvious symptoms.

  2. alanlevinedc Says:

    The authors are correct that ineffective hospital peer review is a problem (Item 5 above). When Lew and i were at OIG, staff from the National Practitioner Data Bank (NPDB) asked OIG to look at the serious problem of hospital under-reporting to the NPDB. Hospitals are required to report clinical privilege restrictions of over 30 days for performance or conduct reasons. In over 20 years, almost 50 percent of the hospitals in the U.S. have never reported a clinical privilege action to the NPDB. For information on the OIG effort, and additional failed efforts, go to this Public Citizen report:

  3. Thomas Cox Says:

    While nobody is likely to dispute that your questions deserve answers across the health care system, they do fall considerably short of the questions that ought to be getting asked and answered.

    Since you are defense attorneys let me fill you in on one of the greatest threats your clients will be facing in the future.

    We are some 4 decades into one of the grandest and least successful social experiments in history. For the last four decades we have increasingly torn apart the fee for service payment system that by all accounts worked quite well for a long period of time and which seems to work quite well in every domain except health care.

    In place of the simple concept of paying for services rendered, we have experimented with a variety of capitation-like health care finance mechanisms: Capitation, Diagnosis Related Groups, Profit/Risk sharing arrangements between insurers and health care providers, the Medicare/Medicaid Prospective Payment Systems for hospitals, physicians, nursing homes and home health agencies. New perturbations include bundled payment systems and episode based care payment protocols.

    The common element in all these ostensibly different approaches to paying for health care services are that they transfer the insurance risks associated with patient care from insurers, managed care organizations and Federal/State programs to health care providers. I refer to this process as “Professional Caregiver Insurance Risk”.

    Setting aside the non-disclosure issues involved when health care providers, including your clients, fail to disclose these relationships to their patients and their patients’ families, a real root cause analysis will answer questions such as the ones I pose below.

    Of course, asking questions is easier than answering questions, so to help you prepare for the coming wave of litigation, I will even answer these questions. If the plaintiff’s attorneys your are dealing with aren’t asking these questions at the moment, they will be doing so in the near future, so you would serve yourselves, and your clients well by beginning to address these issues and your, and their, strategies for dealing with them.

    Here are my deeper root cause analysis questions:

    1. When health care providers accept insurance risks, how is their likelihood of attaining reasonable profit goals affected?

    2. When accepting insurance risks, how likely are health care providers to “break-even” on such contracts?

    3. When they become their patients’ insurers, how likely is it that these risk assuming health care providers will incur operating losses, potentially even insolvency?

    4. As insurers, how does the need to avoid losses and meet profit goals impact the level of care insurance risk assuming health care providers can deliver during fixed financial periods?

    To begin to answer theses deeper questions we need to pull back the Wizard’s curtain and ask the little boy whether he thinks the Capitation-like Emperor is wearing clothes?

    So, to avoid one of the most obvious pitfalls, let us not assume that the health care system is inefficient and that merely becoming more efficient will eliminate any problems the Emperors lack of clothing may cause.

    Instead, let us ask how insurance risk transfers work in efficient health care finance systems (Hint: Not at all).

    In an efficient health care finance system providers specialize in particular types of care. They develop expertise and efficiency through specialization. Neurosurgeons do neurosurgery, pediatricians treat childhood diseases, geriatricians treat the elderly.

    One aspect of our current health care system, exacerbated by the use of capitation-like mechanism, is a de-emphasis on skill and efficiency as providers stretch their skills to treat the broader needs of patient rosters that are not as narrowly defined as the incoming materials on a factory floor.

    This finance mechanism induced inefficiency is the tip of the Titanic sinking iceberg that health care providers, including your clients, face.

    In efficient health care (finance) systems, profit margins for insurers and health care providers are low. Let’s assume 5% on revenues in both cases. When an insurer has a loss ratio of 75% of its premiums, we will assume that it earns profits of 10%.

    But this is insurance, so these profits are not guaranteed. In fact, that 10% is made up of two components: A profit margin of 5% which expresses the target profit goal of the insurer; and a Risk Premium of 5% which is an amount the insurer adds to its premiums as a hedge against failing to meet its profit goal.

    Let’s assume the insurer has 1,000,000 policyholders and that in about 95 years out of 100, this insurer (or about 95 out of 100 insurers with 1,000,000 policyholders in a single year) have loss ratios that vary between 65-85% of premium revenues.

    Since the heart and soul of insurance is the Central Limit Theorem we have a good deal of insight into the implications of these simple assumptions. Since the loss ratios are the average of many, many policyholders we can safely assume that the insurers’ loss ratios are normally distributed. We assume that the mean of this normal distribution is 0.7500 (Loss ratio = 75% of premium revenues) and that the standard error is 0.05, or about 5% of premium revenues.

    Based on these assumptions, these large insurers will have loss ratios lower than 0.7500 about half the time (Probability loss ratio less than 0.7500 = 0.5000).

    The insurer will meet, or exceed, its profit goal of 5% of revenues whenever its loss ratio is lower than 0.8000 (Probability about 0.8413 at a loss ratio = 0.7500 + 0.0500, one standard error above the average loss ratio).

    The insurer will break even, or earn some profits, whenever its loss ratio is lower than 0.8500 (Probability about 0.9772 at a loss ratio = 0.7500 + 0.1000, two standard errors above the average).

    The insurer will lose money whenever its loss ratio exceeds two standard errors above the mean with probability of about 0.0228, at loss ratios greater than 0.8500.

    We are half way to the finish line, but the ride ahead is a lot scarier than the ride thus far.

    Suppose that this insurer decides to transfer its insurance risks to 100 smaller insurers, each of whom accepts 10,000 policyholders, a surrogate for insurance risk assuming health care providers such as your clients. Once again, invoking the Central Limit Theorem, we can specify the standard error of these smaller insurers. The standard error of such a smaller insurer is 10 times larger than the standard error of the larger insurers: 0.5000 as opposed to 0.0500).

    Where the larger insurer has loss ratios that fall between 0.6500 and 0.8500 about 95 years out of 100 (Or about 95 out of 100 such larger insurers can expect their loss ratios, in a single year, to fall between 0.6500 and 0.8500) this is not at all true for these smaller insurers.

    Since these smaller insurers have standard errors of 0.5000, not 0.0500, each such smaller insurer can expect its loss ratio to fall between 0.2500 and 1.2500 about 95 years out of 100 (Or about 95 out of 100 such smaller insurers can expect their loss ratios in a single year to fall between 0.2500 and 1.2500).

    How likely is it that these smaller insurers will earn profits of 10%? Good news! They are every bit as likely to earn profits greater than 10% as the larger insurers. But this is the last of the good news for smaller insurers and the coming wave of litigation facing your clients in the future.

    How likely is it that these smaller insurers will earn profits of at least 5%? Remember, the larger insurers probability of earning profits of at least 5% was 0.8413.

    Bad news! The smaller insurers are nowhere near as likely to earn profits greater than 5% as the larger insurers.

    The problem is that they have very different loss ratio distributions. They have a lot of policyholders, so it is still reasonable to assume that their loss ratios are normally distributed. But their standard errors are markedly different. Instead of Normal(0.7500, 0.0500) like the large insurer, the smaller insurer’s loss ratio distribution is Normal(0.7500, 0.5000).

    The probability that the smaller insurer earns a profit greater than 5% is just 0.5398. Almost 46 out of 100 smaller insurers will fail to earn profits of at least 5% due solely to their small portfolio sizes and their lower probabilities of incurring loss ratios less than 0.8000.

    How likely are these smaller insurer surrogates for insurance risk assuming health care providers to break even? More bad news.

    The probability that the smaller insurer earns some profits, or at least avoids losses, is only 0.5793. Almost 42 out of 100 smaller insurers will fail to break even, incurring operating losses at loss ratios greater than 0.8500 due solely to their small portfolio sizes and their significantly lower probabilities of incurring loss ratios less than 0.8500. Recall that the larger insurer’s probability of incurring loss ratios lower than 0.8500 was 0.9772.

    The larger insurers had virtually no chance of incurring losses in excess of 10% of their premium revenues, at loss ratios higher than 0.9500, not so the smaller insurers. The smaller insurers’ probabilities of incurring operating losses greater than 10% of revenues are 0.3446, at loss ratios exceeding 0.9500.

    In fact, these smaller insurers have the same probability of incurring operating losses greater than 40% of their premium revenues, at loss ratios greater than 1.2500, as the larger insurers have of not breaking even at loss ratios of 0.8500, or 0.0228.

    I have now given answers for the first three questions above and these answers place your clients, and their patients, in considerable jeopardy.

    What, if anything, can these small insurers do to match the ability of the larger insurer to achieve their profit goals of 5% of revenues? Here it is critical to recall that we are talking about an efficient health care (finance) system. The usual answer provided by capitation-like health care finance mechanism advocates: “The providers will be compelled to become more efficient,” will not work. The only way efficient providers can reduce their risk of poor operating results is by cutting, even slashing, patient care costs.

    If smaller insurers, risk assuming health care provider surrogates, do not take immediate steps to cut costs, they cannot alter the above probabilities of missed profit goals and incurred operating losses. If they want to alter these probabilities they must limit the care they provide so that even if they have a bad year, their end of year costs (loss ratios) will not exceed 0.8500.

    Exactly how much they must cut costs by depends on their profit goals and is described in my book: “Standard Errors: Our Failing Health Care (Finance) Systems And How To Fix Them” available at It is too involved for this space.

    What I will say is that if these smaller insurers do not slash anticipated costs by about 60% they are unlikely to meet their goal of earning profits of at least 5% with the same probability available to an insurer with a portfolio 100 times as large. This is the entire basis of how insurance really works and the reason that capitation-like health care finance mechanisms could never work in efficient health care systems.

    When you are already efficient, how do you make such large cuts in service delivery costs? Simple.

    1. Hire less experienced and less skilled staff.

    2. Substitute less trained and lesser skilled patient care assistants for registered nurses.

    3. Load the already pared staff with more patients.

    4. Use cheaper medications than appropriate.

    5. Run operating rooms on an assembly line, leaving little room for unexpected outcomes.

    6. Constantly move patients between floors and nursing units, preventing the formation of the type of bonding nurses are accustomed to and patients have a right to expect when they pay for nursing care.

    7. Constantly tell physicians and nurses that they, not the people who signed off on accepting these capitation-like health care finance mechanisms (Attorneys, accountants and finance directors) are inefficient and responsible for the poor financial outcomes.

    8. Restrict the availability of critically important medications, equipment and supplies.

    I could go on at great length…

    The truth is that the deck is stacked against quality care as soon as health care providers become their patients’ small, terribly inefficient, health insurers. There is absolutely nothing that nurses and physicians can do, in an efficient health care system, that will allow them to assume insurance risks without facing the adverse consequences described above.

    More to the point for you, and for your clients, there is really no defense for knowingly engaging in, or allowing your clients to engage in such activities, because as Paul Harvey used to say:

    “… now you know the rest of the story.”

Leave a Reply

Comment moderation is in use. Please do not submit your comment twice -- it will appear shortly.

Authors: Click here to submit a post.