What Killed Joan Rivers? Piecing Together a Medical Mystery

Screen Shot 2014-09-08 at 10.27.18 AM

There are minor operations and procedures, but there are no minor anesthetics.  This could turn out to be the one lesson learned from the ongoing investigation into the death of comedian Joan Rivers.

Ms. Rivers’ funeral was held yesterday, September 7.  Like so many of her fans, I appreciated her quick wit as she entertained us for decades, poking fun at herself and skewering the fashion choices of the rich and famous.  She earned her success with hard work and keen intelligence–she was, after all, a Phi Beta Kappa graduate of Barnard College.  Ms. Rivers was still going strong at 81 when she walked into an outpatient center for what should have been a quick procedure.

So when she suffered cardiac arrest on August 28, and died a week later, we all wondered what happened.  I have no access to any inside information, and the only people who know are those who were present at the time.

But the facts as they’ve been reported in the press don’t fully make sense, and they raise a number of questions.

What procedure was done?

Early reports stated that Ms. Rivers underwent a procedure involving her vocal cords.  A close friend, Jay Redack, told reporters at the NY Post, “Her throat was bothering her for a long time. Her voice was getting more raspy, if that was possible.”  In a televised interview, Redack told CNN that Ms. Rivers was scheduled to undergo a procedure “on either her vocal cords or her throat.”

However, the Manhattan clinic where Ms. Rivers was treated, Yorkville Endoscopy, offers only procedures to diagnose problems of the digestive tract.  All the physicians listed on the staff are specialists in gastroenterology.  Any procedure on the vocal cords typically would be done by an otolaryngologist, who specializes in disorders of the ear, nose, and throat.

So it may be that acid reflux was considered as a possible cause of Ms. Rivers’ increasingly raspy voice, and she may have been scheduled for endoscopy at the Yorkville clinic to examine the lining of her esophagus and stomach.  Endoscopy could reveal signs of inflammation and support a diagnosis of acid reflux.

Upper gastrointestinal (GI) endoscopy involves insertion of a large scope through the patient’s mouth into the esophagus, and passage of the scope into the stomach and the beginning of the small intestine.  It’s a simple procedure, but uncomfortable enough that most patients are given sedation or, less commonly, general anesthesia.

Was sedation given?

Three types of medication are commonly used for sedation during endoscopy:

1.  Midazolam, diazepam (Valium), or other medications in the benzodiazepine family are often used to help patients relax before the start of the procedure and to produce amnesia.

2.  Narcotics such as Demerol and morphine are often used to provide pain relief and make the procedure less uncomfortable.

3.  Propofol, a potent sedative and hypnotic medication, may be used to induce sleep and prevent awareness.  Many people first heard of propofol as the medication associated with the death of singer Michael Jackson in 2009.

Any of these medications may lower a patient’s blood pressure and depress breathing.  Continuous monitoring of vital signs by trained personnel is critical, both during the procedure and in the recovery room.  Breathing may be especially difficult during upper GI endoscopy because the scope may partially block the patient’s airway.

There is no public information yet about exactly what procedure Ms. Rivers underwent, what type of sedation she may have received, how her vital signs were monitored, or who was in charge of administering the sedation. Typically, benzodiazepines and narcotics may be administered for endoscopy by registered nurses (RNs) who are supervised by the gastroenterologist who performs the procedure.

Was propofol used?

Controversy exists over whether or not safe propofol use requires a higher level of training than RNs receive.  The package insert for propofol clearly states:

“For general anesthesia or monitored anesthesia care (MAC) sedation, propofol injectable emulsion should be administered only by persons trained in the administration of general anesthesia and not involved in the conduct of the surgical/diagnostic procedure. Sedated patients should be continuously monitored, and facilities for maintenance of a patent airway, providing artificial ventilation, administering supplemental oxygen, and instituting cardiovascular resuscitation must be immediately available. Patients should be continuously monitored for early signs of hypotension [low blood pressure], apnea [nonbreathing], airway obstruction, and/or oxygen desaturation [lack of oxygen].”

The package insert goes on to emphasize that adverse effects of propofol are more likely to occur in elderly and frail patients.

Many anesthesiologists (myself included) use propofol often because patients wake up quickly afterward, and it causes less nausea than other sedatives.  But propofol can be fatally easy to use, and there is no absolute safe dose. Sometimes even a small extra amount is enough to make breathing slow down or stop.

Despite these warnings, the American Society for Gastrointestinal Endoscopy (ASGE) maintains that propofol may be safely administered by a nurse under the supervision of a gastroenterologist, and that the expertise of a physician anesthesiologist isn’t necessary for the care of healthy, low-risk patients.

Were there other risk factors?

Would the 81-year-old Ms. Rivers have been considered a healthy, low-risk patient?  We know little about her health.  But as far back as 1985, she told the audience of Good Morning America that she occasionally experienced “arrhythmia” (irregular heart beats), which “scares the hell out of me.”

We do know this much:  on the morning of August 28, emergency medical responders were called to Yorkville Endoscopy where they found Ms. Rivers “unconscious and in cardiac arrest“, the New York Times reported.  She was taken to Mount Sinai Hospital, where for several days she remained in a medically-induced coma.  This is a technique typically used in cases of brain injury, with the hope of reducing pressure inside the brain and temporarily reducing the brain’s activity and oxygen need.  Once the coma-inducing medications are stopped, the patient is observed for signs of brain recovery.

On September 2, Melissa Rivers reported on her mother’s website that that she remained on life support.  The next day, though, Ms. Rivers was “moved out of intensive care and into a private room where she is being kept comfortable.” On September 4, Melissa announced that Ms. Rivers “passed peacefully at 1:17 pm surrounded by family and close friends.”

The move out of intensive care probably reflected the fact that recovery was not expected, and comfort measures were the only treatment planned.  From this information, it appears likely that Ms. Rivers was successfully resuscitated from the initial cardiac arrest and that her heart resumed beating.  However, her system suffered irreversible damage.

Pushing the limits of outpatient care

There is no way to know, without further information, whether the root cause was trouble with her heart, her breathing, a sudden stroke, or another type of catastrophic event.  There is no way to know if problems were due to sedative drugs she might have received.

There is no way to know, without further information, if the extra equipment and personnel available in a full-service hospital as opposed to an outpatient clinic would have made any difference in Ms. Rivers’ resuscitation and outcome.

But this much is clear:  there is pressure today from the government and insurers for physicians to perform complex procedures even on high-risk patients in free-standing ambulatory centers.  Why?  To save money.  The extra equipment and staff in full-service hospitals are expensive.  Moving procedures to streamlined outpatient settings is cheaper, and patients find the environment more pleasant.

There are still risks in every medical procedure, and the risks are higher for older patients with underlying health problems.  Even if a patient has undergone surgery and anesthesia on previous occasions without problems, complications may occur.

I’m certain that the physicians and staff at Yorkville Endoscopy are devastated by the events of August 28, and I’m equally certain that every aspect of what happened will be reviewed in microscopic detail.  But it may be that we’ve pushed outpatient care as far as it should go.  We need to acknowledge that invasive procedures are just that–invasive–and that the medications used in sedation and anesthesia can be deadly when we least expect it.

33 replies »

  1. whoah this weblog is wonderful i really like reading your posts.
    Keep up the good work! You realize, many people are
    searching around for this info, you could help them greatly.

  2. I have the image in my mind of a physician who had (in common parlance) ‘gotten too big for his britches’ psychologically. Caution, conscientiousness and concern are the codewords. Arrogance is equal to ignorance. I wonder if the Endoscopy Center had any concerns about this doctor prior to this. Were there any prior ‘close calls’? Why did this happen specifically to Joan? Will we ever be able to find out?

  3. The question I’d like to know the answer to is this: If Joan Rivers experienced laryngospasm, was it treated in an appropriate and timely fashion? Most anesthesiologists are extremely adept at treating laryngospasm. Even refractory cases can be treated immediately with succinylcholine. Was succinylcholine immediately available? Was it given? Also, if hypoxia led to cardiac arrest and subsequent brain damage, why wasn’t a surgical airway (emergency cricothyroidotomy) performed? I’m just asking based on the “facts” the media pundits have laid out.

  4. Curiouser & curiouser.

    It’s been a *long* time since I was involved in a code, but found myself wondering how they sorted out who would take charge of the code…presumably when the firefighters & paramedics respond to a 911 call, they don’t usually find a squad of doctors working on an emergently intubated patient.

    Also, I was struck by the volume at the clinic: “the clinic had performed 18,000 procedures since it opened in February 2013.” That means 45 procedures per business day…This struck me as a lot, but maybe this is common in endoscopy clinics?

  5. Updated information from the New York Times–

    The Times reports that Dr. Leonard Cohen, a gastroenterologist, was Ms. Rivers’ physician at Yorkville Endoscopy. As PJM noted above, Dr. Cohen has been a prominent supporter of allowing gastroenterologists to supervise nurses administering propofol, without any involvement or presence of a physician anesthesiologist or any other qualified anesthesia practitioner. It’s still unclear who was monitoring or giving sedation to Ms. Rivers at the time.

    Speculation still abounds. See also:

  6. The total costs at ambulatory centers are lower than hospitals. But remember these centers are not required to treat the uninsured whereas hospitals are and they are a big loss for the hospital.

  7. Thank you, Dr. Sibert. I have been wondering about this for days and looking for answers about what possibly could have gone wrong. I agree that Joan was in the midst of a busy schedule and her daughter was not even in town, so I’m assuming that she was assured that nothing dangerous was involved. Regardless of the Ms. Rivers’ age, she was very vital and not about to die. So I am interested in what went wrong and this is the best and most informative blog (and responses) that I have found.

  8. Total charge in an Endoscopy Center could include different components.
    Some charges would be: GI doctor with a procedure code for doing the Endoscopy. Another charge for the anesthesiologist. Another separate charge for the facility fee.
    Believe it or not, the allowed amount by the insurance company for any of the component pieces can and does vary depending on the location of the service provided.
    The fee allowed to the GI doctor can be different depending on the location of the service provided (in a hospital OR, in a hospital surgery center, in a privately owned Ambulatory Surgery Center, in an office).
    The facility fee allowed is different depending on where the procedure is done. The allowed amount for the room for doing an endoscopy is absolutely different depending on location: in a doctor’s office, in an “approved” “accredited” ambulatory surgery center, in a hospital operating room. The total charge is higher if done in a hospital operating room.
    And, not all GI doctors typically go to the hospital to do these types of procedures- for various reasons!
    And the charges allowed can vary from institution to institution, even just blocks away from each other.
    There is incentive to patients and doctors to provide services in an ambulatory type center because in a hospital the percentage paid by the insurance company can be lower and therefore the patient could end up paying more out of pocket costs when the same procedure is done in a hospital setting; and that does seem to create pressure for location of doing various procedures.

  9. Barbara Ann,

    Are you sure that the total charge in an Endoscopy Center are the same as those in a Hospital?

    I can assure you that that is NOT the case with outpatient imaging vs. inpatient imaging. Outpatient imaging is substantially cheaper.

  10. Hopefully the facts will eventually come out and be made public. I am a bit disturbed that one of the gastroenterologists listed on staff at the clinic has been a vocal critic of the FDA black box warning regarding propofol. When the FDA rejected a petition from the ACG seeking removal of the warning in 2010 this physician was dismissive of the FDA findings. He suggested that in the endoscopy suite propofol can easily be targeted to moderate sedation, safely given by non-anesthesia providers, and that deep (General) anesthesia is never required. I just wonder if his naive view of propofol’s risks is shared by his colleagues at the clinic.


  11. There is so much difficulty in defining a “general anesthetic” vs. sedation. When propofol is used, the patient is asleep. That can either be defined as deep sedation or as general anesthesia without a controlled airway. The difference is semantics. Either way, the patient needs a great deal of help to make sure that a patent airway is maintained, and that vital signs are stable.

    And that’s the whole problem with propofol. Sedation looks simple and pleasant until it goes wrong.

    General anesthesia usually (though not always) implies that the airway is controlled by the person administering the anesthesia. This can be with an endotracheal tube, an LMA, or simply a mask. Sedation, on the other hand, implies that the patient is still in control of his or her own breathing and airway. This can be a lethal assumption, depending on the depth of sedation.

  12. I don’t think Joan would have been having an operation of any kind as she was scheduled to perform the next day, was in the work-up to NY Fashion Week, the biggest event of the year, and Melissa was in LA. So, presumably it was a diagnostic procedure. Would that normally require a general anesthetic?

  13. To Dr. Sibert,
    Your blog article was excellent; so informative and helpful. When I wrote a response about “some confusion” I was not directing it at you! I have read so many comments on various sites I just felt I would clarify that one point.

    Your reply does concern me about larger type procedures. If they technically can be performed in an ASC type facility as out-patient, then patients are going home soon after surgery. The level of pain and other issues can present a problem. I unfortunately know about this first hand.

    I do feel that whatever occurred regarding Joan Rivers will be helpful to know and understand so that any future patient can be helped. I look forward to your comments when more information is known.

  14. The article has an incorrect fact.There is no “pressure” by the government to have endoscopy procedures done in an ambulatory care setting as a way to cost save! There is a higher profit margin for the owner of these ambulatory care settings. Many Gastroenterologists have a stake thus patients are encouraged to have their gi procedures done in their ambulatory care center. Hospitals offer the same services and are highly regulated. They have credentialed nurses/techs/labs/ code teams as well as state of the art equipment and meds such as ventilators, epinephrine, sux What does the ambulatory care center have???

  15. This seems to be a sad story. As a geriatrician who always brings up advance care planning with patients, I find myself wondering what kind of health crisis planning Joan Rivers had done prior to all this.

    I hope she had an advance directive and that this provided some guidance to her clinicians and daughter during those terrible days when she was in the ICU.

    BTW I don’t like to refer to this as “end-of-life” planning even though that’s catchy. The end-of-life is easy to see in hindsight and often harder to spot in real-time, so I usually call this planning for health crises and possibly-end-of-life.

  16. Yikes1–you’re certainly correct about the possibility of an ENT surgeon coming to do a VC procedure. It’s all speculation, but nonetheless the sedation and airway issues wouldn’t be very different.

    You’re also correct about the definition of “outpatient”. That’s why I specifically mentioned the pressure to do procedures at free-standing ambulatory centers.

    High risk patients are often best taken care of in the hospital for their procedures even if they do well enough to go home the same day, which is of course what everyone hopes. Whenever I go to an ASC, I always have the uncomfortable feeling that I’m working in a satellite orbiting the mother ship–close, but not close enough to get real help if it’s needed. In California, for example, you can’t transfuse blood in an ASC–no matter how desperate the situation, the pt has to be transferred to a hospital first. And the trend is to do ever larger procedures–robotic prostatectomies and spine surgeries are just the latest examples.

  17. Why speculate about anything? Besides in this case “review” is a legal tactic.

  18. It is entirely possible that an otolaryngologist (or another type of doctor) was given permission/privileges to perform a procedure on an as needed basis and not required to be part of regular staff. The major reason to use this type of center is because of the ability to have anesthesia (beyond local anesthesia) for one’s patient, in an out-patient setting.
    It is also possible an outpatient surgery center setting was discussed and chosen because it is more private and less chance than as in a hospital of people knowing that a procedure was being done on a specific person.
    I think there is some confusion: An out-patient procedure means a patient is not admitted into a hospital for 24 hours or more. Even if surgery is done inside a hospital , in an operating room, it is considered out-patient if one does not remain for 24 hours. If Yorkville Endoscopy had the necessary emergency equipment and necessary staff then regardless of where a procedure involving sedation/more than just local is done the outcome should be the same. {i hope that makes sense}.

  19. Excellent summary. And I agree with the comment that no one that age should receive anything but emergency medical treatment, especially involving anesthesia.

  20. Tricky one.

    There’s a compelling narrative here: low unit costs, high margins, payer trying to save costs on old and sick people.

    This narrative is very likely correct but I still want to know the cause of death and the likelihood this would have been averted if procedure was not done as an outpatient.

  21. Excellent post and comment threat. Yet another indication that we’ve built a healthcare system around entrepreneurship and “do more, cuz it pays better” instead of patient safety and effectiveness.

  22. What with all of the great technological advances in surgery and anesthesia, it is easy to become complacent and forget that there are very real risks in these medical procedures.

  23. Excellent analysis.

    The who is allowed to give what thing is a massive big deal.

    This case is really about how we assess risk and complexity.

    How risky of a procedure is this?

    Not risky at all. Unless something goes wrong.

    In which case it is extremely dangerous.

  24. If propofol was administered by a nurse under the direction of the gasteroenterologist (the likely scenario given the setting and the type of procedure), you have a very strong indication of the likely cause ..

    This is why conservative physicians hesitate before taking a “minor” case involving an older patient