A few weeks ago, a middle-aged man decided to tweet about his mother’s illness from her bedside. The tweets went viral and became the subject of a national conversation. The man, of course, was NPR anchorman Scott Simon, and his reflections about his mother’s illness and ultimate death are poignant, insightful, and well worth your time.
Those same days, and unaware of Simon’s real-time reports, I also found myself caring for my hospitalized mother, and I made the same decision – to tweet from the bedside. (As with Simon’s mom, mine didn’t quite understand what Twitter is, but trusted her son that this was a good thing to do.) Being with my mother during a four-day inpatient stay offered a window into how things actually work at my own hospital, where I’ve practiced for three decades, and into the worlds of hospital care and patient safety, my professional passions. In this blog, I’ll take advantage of the absence of a 140-character limit to explore some of the lessons I learned.
First a little background. My mother is a delightful 77-year-old woman who lives with my 83-year-old father in Boca Raton, Florida. She has been generally healthy through her life. Two years ago, a lung nodule being followed on serial CT scans was diagnosed as cancer, and she underwent a right lower lobectomy, which left her mildly short of breath but with a reasonably good prognosis. In her left lower lung is another small nodule; it too is now is being followed with serial scans. While that remaining nodule may yet prove cancerous, it does not light up on PET scan nor has it grown in a year. So we’re continuing to track it, with crossed fingers.
Unfortunately, after a challenging recovery from her lung surgery, about a year ago Mom developed a small bowel obstruction (SBO). For those of you who aren’t clinical, this is one of life’s most painful events: the bowel, blocked, begins to swell as its contents back up, eventually leading to intractable nausea and vomiting, and excruciating pain. Bowel obstruction is rare in a “virgin” abdomen – the vast majority of cases result from scar tissue (“adhesions”) that formed after prior surgery. In my mother’s case, of course, we worried that the SBO was a result of metastatic lung cancer, but the investigation showed only scar tissue, probably from a hysterectomy done decades earlier.
The treatment for bowel obstruction is to place the bowel at rest (“NPO”) and to suction out the contents of the stomach through a nasogastric (NG) tube, which decompresses the GI tract and decreases the pain and distension. And then you wait. Many SBOs will open up on their own, but those that don’t after a reasonable waiting period (4 to 7 days, generally) require surgery. Mom’s first obstruction didn’t open up, and a mercifully uncomplicated laparoscopic procedure successfully cut the fibrous band that was kinking the bowel (“lysing the adhesion”), relieving the obstruction.
Unfortunately, she had two more cases of SBO in the next year; these resolved after 3 to 5 days of bowel rest without surgery. We hoped we had seen the end of it, but we hadn’t: on a visit last month to the Bay Area to see my sister and me, my mother developed another SBO, a few hours before getting on the plane to Florida. Her pain began slowly, but within an hour she was in agony, retching, sweating and literally shaking. (By the way, nothing prepares adult children to see their parents in this much distress; it seems so very wrong, the world turned on its head. One of Scott Simon’s tweets captured my feelings: “I wish I could lift my mother’s pain & fears from her bones into mine.”) I took her to UCSF Medical Center’s emergency department. The next day, I sent my father – who is becoming a bit frail and whose presence by the bedside of his wife of 59 years was likely only to make everyone more anxious – back to Florida.
I’ve “eaten at my own restaurant” many times before: my two sons were born there, I’ve had surgery there, and I’ve accompanied family and friends to a dozen or so ED visits over the years. But being there with Mom somehow felt different: first of all, it was my mother; second of all, her hospitalization was likely to last several days (culminating in either resolution of the SBO or surgery), with brief periods of excitement punctuating long periods of waiting. This offered me the chance to soak in the processes and the culture. Here are six lessons I drew from my observations:
Chokepoints: One of the key principles in “Lean thinking” is to identify steps in a care process that don’t add value. Within a few minutes in the ED, I found a doozy. There was little question that Mom was having another bowel obstruction and, just as clearly, was going to end up with an abdominal CT scan. The scan has become trivially easy to do: there’s a scanner in the ED, and the test now takes about 10 minutes to complete. Blood tests were drawn and some results, such as the CBC, returned quickly. A few minutes after the blood was sent to the lab, the radiology tech arrived to obtain a chest x-ray and KUB (two different x-rays). Why did she need these, I wondered, when she’s going to get a CT scan in a few minutes?
It turns out that even a “stat” creatinine takes an hour or so to return, and radiology, understandably, won’t do a CT scan with contrast in the absence of a recent test of kidney function. Given the small chance that Mom was suffering from some other abdominal catastrophe like a perforation, the ED docs felt that an hour’s wait for the CT was too long. And so came the stat KUB/CXR – at a cost of several hundred, perhaps even one thousand, dollars – all because the creatinine was not more promptly available. It made me wonder: has anybody done the accounting to figure out the cost of a more rapid creatinine assay, factoring in the savings (in dollars, motion, and radiation) from obviating the need for the other, inferior, studies? And how many other chokepoints like this are scattered around our system?
Introductions and Rounding: On our medical service at UCSF, we’ve worked hard at improving the way we introduce ourselves, and my experience with Mom made me glad of this. We now have a white board by each bedside, on which all of the players (attendings, residents, students, nurses) are supposed to write their names, although adherence is less than perfect. Moreover, all our physicians now hand patients “face cards” with our pictures, names, titles, education, and a description of our role.
On Mom’s surgical floor, while the nurses and patient care assistants (PCAs) were excellent at introducing themselves and writing their names on the whiteboard, the doctors were not. Some didn’t introduce themselves at all; most did, but their roles were unclear. There were no face cards, and no physician names on the whiteboard. While the physicians in the ED did a good job introducing themselves, once we made it to the surgical ward, it was more than six hours before we had any idea who her primary doctor or team were. Later, when they came in to see her, they left behind no written traces.
We need to standardize, and then modernize, this effort. Sure, let’s start with cards and whiteboards. But how about this for an ideal futurist scenario: when doctors or nurses enter a room, our badge automatically triggers our picture, bio, and position description to pop up on the patient’s television screen or tablet computer, all recorded so that they can be shared with family members and looked up later.
Which brings me to rounding practices. On UCSF’s pediatrics inpatient service, teams now prospectively plan their morning rounds and – are you sitting down – produce a schedule that tells patients and families when the team will be by to see them. This schedule allows the nurse to be there, as well as some of the relevant subspecialists. We’re currently trying to figure out if we can pull this off on our much bigger and more geographically scattered medical service. I will push for us to do so, particularly after my experience with my mother.
I knew the surgeons would be there at some point between 6:30am and 9am, but had no way of pinning down the time beyond that. (Think about how you feel when Comcast says that the cable guy will be there between 1pm and 5pm.) I also knew that once the surgical team members left the floor, I would have no further access to them for the rest of the day. Most mornings, I came in by 6:30 and waited. One morning, I got there five minutes late, but luckily caught them before they left the ward. And one morning, I missed them and, as Murphy would have it, that was the occasion for the worst error of the hospitalization.
(Far) Too Much Information: On the morning I missed the surgeons on rounds, I had given myself a treat, stopping at Starbucks on the way in to the hospital. While sipping my mocha, I received a frantic call from my dad in Florida. “One of the doctors was just in to see your mom, and told her she has lung cancer,” he said, his voice shaking. I’m on my way in, I told him.
You can guess what happened. A harried July intern was running from room to room seeing patients, her mind probably distracted by the thoughts of everything she needed to get done that day. She saw the official reading of the abdominal CT scan from the prior day. On it, Mom’s left lung nodule was seen and called “suspicious for malignancy.” (Her prior films are in Florida; the UCSF radiologists had nothing to compare this to.) Rather than calling me about the finding (yes, everyone knew who I was [though I remain unconvinced that VIP status necessarily improves one’s outcomes] and my cellphone number was on the whiteboard), the intern came into Mom’s room, woke her from her morphine-aided sleep, and told her that the x-ray showed probable lung cancer. My mother wasn’t sure if she had dreamed this. Unfortunately, she hadn’t.
I was livid, but now have the distance to draw some general lessons from this error:
- The word “cancer” remains emotionally nuclear; it should not be tossed around lightly.
- Don’t give weighty, complex, and highly charged information to narcotized elderly people – or anyone for that matter – who are just waking up without having their family there.
- The intern should have made a little effort to inquire about prior films.
- We really need interoperable health records, which would have allowed our radiologists to access Mom’s Florida scans.
- Even when you’re rushing through your checklist, please use a little common sense.
The Nurses: I’ll not be the first person to observe that it is the nurses, not the doctors, who really determine the overall experience of a hospitalization. Our nurses, on 13 Long at UCSF Medical Center, were exceptional. The nurses and the Patient Care Assistants (PCAs) were competent and caring, and they had their acts together. In the old days, our nurses wore varying types of clothes, the roles of nurses and PCAs were vague, and basic processes like medication administration and shift change were done haphazardly. Today, UCSF nurses are dressed professionally, in uniforms that make clear their roles; their handoffs are robust (“I’m signing off in 30 minutes; this is Anna, who will be taking over for me”); it’s clear what the nurses do and what the PCAs do (there is a special place in heaven for the PCAs, by the way); and the presence of bar coding and computers in the rooms makes things feel far safer and has standardized previously chaotic processes. (All of these elements are captured by our achievement of “Magnet Status.”) One nurse tried an IV twice, couldn’t get it, and calmly said, “I wasn’t successful; I’ll get my colleague who will get it done.” (She did.) And everybody cleaned his or her hands, every time. It was very impressive.
Part of the reason that the experience with the nurses is so important is that it either does – or doesn’t – give you as a family member trust that things will be okay when you’re not there. When Mom was hospitalized in Florida last year, my father hired a private nurse to be with her one night; on another, I asked to sleep at her bedside, and was given a cot so disgusting that it would have been rejected by every youth hostel I stayed in during college. At UCSF Medical Center, I left at 10pm each night with complete confidence that things would be okay overnight.
Yet there are nursing-related processes that can be improved. Two bits of technology remain the bane of patients’ lives in the hospital: the nurse call button and the IV pump alarm. In 2013, in a hospital in San Francisco, can we really not come up with a system where the call button triggers a video feed to allow a nurse to see and speak to the patient, allowing for effective triage? And the IV pump beeps incessantly: kinked line, empty IV bag, incoming tweet, who the hell knows. Isn’t there some new technology to improve this situation?
(Speaking of new technology, the NG tube also seems ripe for innovation. Rigid and uncomfortable; it gives poignancy to the old schoolyard insult, “up your nose with a rubber hose.” Can’t we do better?)
Jargon: Boy, this one is tough. I’ve got a pretty good jargon detector, but after watching the docs speak to Mom for a few days, I realized that half of what I say is probably unintelligible to my patients. One of the surgeons, who I regard as a fabulous communicator, came in one day to talk about Mom’s surgical options. “We can go in there and try to lyse the adhesions, but this only will work if we can find a focal lesion,” he said. Not the worst I’ve heard, but listening through my mother’s ears, I realized that she did not understand anything after, ‘We can go in there…” On our rounds, perhaps we should appoint one team member to be our official jargon-detector, offering the speaker feedback in an effort to sensitive all of us to this pernicious problem. (Sure, the patient-centered view would have us empower patients to speak up if they don’t understand something, but I think few patients will. Certainly not my mom – unfailingly polite, she just shook her head in agreement, despite understanding very little).
Managing Uncertainty: Because of my VIP status, we had a lot of doctors weighing in on the best treatment. The answers to some questions were clear, including the desirability of waiting for spontaneous resolution rather than doing urgent surgery. Thankfully, on Day Three her bowel opened up and she began to eat. (By the way, the only wonderful thing about NG tubes is that taking them out is one of life’s great pleasures; think ski boots and multiply by 100.) But whether she should go for elective surgery, given the near-certainty of future recurrence, and whether and how to modify her diet, were areas of great ambiguity.
On each of these, I liked the fact that the physicians laid out the plusses and minuses of the various actions. What was difficult – even for me – was how to deal with conflicting recommendations. Many folks emphasize the importance of a strict low-fiber diet, and yet one surgeon made clear that it probably doesn’t make any difference. (Mom chose to stick with the diet; mainly because doing something feels better than doing nothing). And we decided against elective laparotomy to “clean out the abdomen,” a big operation with no guarantee of success. This was the consensus judgment of several physicians, although there were a number of variations on the theme that could have led to some confusion.
All of this made me more convinced than ever of the need for a quarterback, one physician to synthesize all of the conflicting advice and present a unified recommendation to the patient and family. It’s what a good hospitalist does. In my mother’s case, this role was assumed by the lead surgeon, who also supplied me with some of the key literature, which was extremely helpful.
Unlike Scott Simon’s mom, my mother’s hospitalization had a happy ending, at least for now. Overall, I was very pleased with the quality of care she received at UCSF Medical Center and would recommend it to other patients without the slightest hesitation. The nurses and doctors we met are bright, committed, friendly, and collaborative, and the systems we’ve put in place over the past decade – IT, handoff processes, whiteboards, and more – are making a difference.
That said, like all great organizations, every encounter provides an opportunity for improvement. I hope we can use my mom’s experiences – and those of other patients and families – to do a little better for the next patient.
Robert Wachter, MD, professor of medicine at UCSF, is widely regarded as a leading figure in the patient safety and quality movements. He edits the federal government’s two leading safety websites, and the second edition of his book, “Understanding Patient Safety,” was recently published by McGraw-Hill. In addition, he coined the term “hospitalist” in an influential 1996 essay in The New England Journal of Medicine and is the former chair of the American Board of Internal Medicine. His posts appear semi-regularly on THCB and on his own blog, Wachter’s World.
Great post. It is amazing what we can learn by looking in from the “other side.” I tend to be amazed at how much of the patient experience most medical professionals don’t understand and never think about.
The next step is improving the process and connections beyond the hospital walls. My parents are meeting with or just finished a meeting with mom’s surgeon as I typed. Last week when I spoke with the surgeon, I asked if I could be dialed into the discussion. He didn’t have a speaker phone. I asked if my parents brought their iPhones could he use that speaker phone. The response “I’ve never done a consult like that. I don’t know how I would make it work.”
Didn’t doctors play the telephone game as kids? What don’t you understand about the fact that not having everyone hear the same thing at the same time decreases the accuracy of the information exchange?
I’ve spent too much time inside a hospital – first with my mom’s never-ending cancer treatments and now with my father’s health. I’ve seen good docs and bad, but the best staff was at Wake Forest. Everyone there made my family feel as best as we could, and we had complete trust in their abilities.
Very nice insight on what it feels like inside the hospital. I too have had way too many visits as my mom is suffering from kidney disease. It’s difficult yet great to see that there are people, doctors, nurses and coordinators who go out of their way to give you extra care when needed.
Great post and thanks for sharing your mom’s story. I agree that It is certainly very revealing to get care at your own healthcare facility. With respect to the discussion about July, it is also my observation the best attendings often are on in July and providing extra supervision. Nurses are also on guard. As anyone knows, it is fatiguing to be on in July though since things take longer and more diligence is required to review. One solution that has not been well explored is adjusting workload. Unfortunately, medical intern patient caps (limit on number of patients they care for) are often the same on the first day as the last day of internship, despite knowing that the workload is more of a stress early in their career. One way to solve that is to actually reduce workload for new interns in early part of their internship so they ‘graduate’ up to higher workload caps. While this may make sense educationally, the problem is that from a hospital perspective this does not make sense financially as the work is reassigned to hospitalists. It’s also worth nothing that often new interns have not seen a patient for MONTHS (sometimes more than 6!). This could certainly be a cause of inefficiency when they start their internship. In addition to learning the new hospital, they are relearning how to care for patients efficiently. Imagine if the last rotation in MS4 year was a more intense clinical rotation that ramped up to internship. Im not sure it would be popular, but it would certainly likely make the transition to internship a bit easier. For this reason, many GME programs are now using boot camps to assess and improve core skills, which is very intense intervention compared to GME orientations in the past. There is no easy answer for July, but there are things we can try to make it an easier transition.
We all dread July but I’m not sure it is because we genuinely believe that patients are in danger. It is because of the extra work it takes to orient and supervise the newbies.
Mistakes do happen but the evidence on whether patients are actually harmed more is mixed at best. In my experience, the short leashes and close supervision may actually make patients safer by creating important redundancies (at the cost of efficiency) compared to other times of the year.
Thanks for another fantastic and insightful post Bob.
I can empathize with Bob and Scott. My own experience with my mother’s death in early 2000 was equally jarring. It began with her having a major stroke at 1 am on 1/1/2000. By the time the rescue squad got her to the hospital, she’d suffered the first of two MIs; the second would come about four hours later, in what seemed to me at the time (only child, father already deceased) like my known universe coming apart at the seams. Fortunately, she was soon stabilized although she made it quite clear that she was ready to die, and she did soon thereafter, bringing respite that had eluded her here.
The physicians and nurses who cared for my mom were beyond reproach (with the exception of one overzealous cardiologist who needed her consciousness rearranged, which I was only too happy to do, when it came to the topics of intubation and resuscitation).
But, as I noted in my comments on Ashish Jha’s recent personal hospital experience (https://thehealthcareblog.com/blog/2013/08/14/average-care-at-a-typical-hospital-on-an-ordinary-sunday/), I think we miss an important and potentially unmatched quality improvement opportunity if we don’t encourage people to use social media to put pressure on the health care system. For too long, health care providers have gotten the benefit of the doubt, been given a pass on both poor behavior and poorly rendered care. What if the family of Josie King had been able to raise a social media ruckus about what was happening to their daughter (http://www.josieking.org/Home) at the vaunted Johns Hopkins Hospital in 2001?
In my recent reading experience, social media types in health care institutions want to use social media as a way to “communicate” with patients, extol the virtues of their institutions, and highlight their new products and services. This is all another way of saying social media is the next vehicle for optimizing payer mix and maximizing reimbursement. That’s not good enough.
The reality remains that in an culture where there is deep (and often justified) suspicion of corporate interests (and that’s mostly what a large hospital or health system is), there is no reason to exempt medical care providers from feeling the immediate sting of negative trending. It might be the easiest and most efficient way for US health care consumers to send the message that the system stinks (except for the most savvy), that health reform has made it only slightly less smelly, and that health care providers churn excuses faster than farmers churn milk into butter.
Thanks for an outstanding article.
Scott — thanks for taking the time to read my thoughts and to comment. The evidence that care is worse in July (known in the US as “the July effect”; in the UK as the more macabre “August Killing Season”) is spotty (and there is really no evidence that care is worse in teaching hospitals, overall). Some studies show a definite rise in mortality in July, others don’t. In good teaching hospitals like UCSF, we put our best supervising physicians on and keep our eyes open a little wider, the leash a bit shorter. That said, based on my experience more than the literature, if I could choose (ie for elective surgery), I’d probably wait till later in the year. However, as you know, most inpatient care is not volitional.
The point you raise regarding reimbursement is a subset of a far larger issue in healthcare: should better care be paid better, regardless of the reason (time of year, better staffing, better training)? Here too, alas, the answer is: it’s complicated. The evidence that differential payment makes a difference in healthcare is underwhelming, though it’s certainly being tried, as part of today’s kitchen-sink approach to improving quality, safety, and efficiency. The problems with so-called “pay for performance” are many: healthcare people are motivated by more than money (as are NPR people); we’re not very good at measuring quality; and very large determinants of outcomes have nothing to do with the quality of care. In fact, if we’re not careful, a payment system can easily cause hospitals to shy away from taking poor patients, or those with limited social supports — since they are more likely to have worse outcomes.
One final point that adds yet one more wrinkle. When it comes to training doctors, nobody has come up with a better idea than having trainees actually begin taking care of patients, with supervision. There are ways of doing it better and safer (practicing on simulators, for example), but the bottom line is that there is that we all benefit from having trainees work on real patients. We all want experienced doctors, yet there is no way to get to your 100th case without somehow going through your first.
Thanks again. My best to you and your family — I’m a great admirer of your work and your tweets were really special.
Either that or more people will show up to take advantage of the “summer specials!”
I enjoyed reading through Dr. Wachter’s thoughts, even those that a lack of medical knowledge prevents me from understanding. And I’m happy for his family. There are a lot of good ideas there.
I am also amused by that phrase “July interns.” Do hospitals reduce their billings in July, because of staff inexperience and mistakes? We had some similar instances which, given my mother’s more dire results, are painful to recall now. But necessary.
I have an idea: reduce the costs that are billed in July and August, on the idea that everybody seems to know that the level of care is not as good. This will provide hospitals with an incentive to improve care to earn more money.