Published in the June 2013 issue of Today’s Hospitalist
HOW MANY PATIENTS per day is too many? And given how many factors can affect your ability to provide safe care ” from having to cover ICU patients to being able to depend on midlevel providers “how do hospitalists even define what an ideal workload is?
To answer those questions, a research team has launched an ambitious series of studies. Researchers first reached out in 2010 to more than 500 hospitalists via QuantiaMD.com, an online physician community, to ask how often they felt their census was unsafe and what a safe workload would look like in their particular hospital.
Their findings, published in the March 11, 2013, issue of JAMA Internal Medicine, made headlines. While surveyed hospitalists generally felt they should see no more than 15 patients a day, 40% reported exceeding their own perceived safe level at their institution at least once a month. Even more alarming, 36% of those claimed they had to cope with unsafe workloads more than once a week.
When asked to identify the adverse effects of seeing too many patients, 25% of respondents said that too much volume had prevented them from fully answering patient questions or discussing treatment options within the last year. And 22% said that workload issues had led them to delay admissions or discharges.
The same research team was back the next month, this time with a follow-up study published online by JAMA Internal Medicine in April. The authors interviewed hospitalist program directors and met with small groups of hospitalists, non-physician providers, housestaff and administrators from private, community and academic settings.
Their goal was to tease out specific factors “relating to physicians, hospitals, teams and patients “that affect hospitalist workload. Factors affecting workload ranged from individual physician ages and years in practice to patient acuity and whether or not a hospital had a plan in place to boost physician staffing.
“The second study was to provide a context for the numbers of encounters reported in that first survey,” says lead author Henry J. Michtalik, MD, MPH, a hospitalist at Baltimore’s Johns Hopkins Hospital and associate faculty for the Armstrong Institute for Patient Safety and Quality at Johns Hopkins. “We recognized that many factors affect workload, but we also know that national benchmarks don’t report these. So it is very hard to compare apples to apples when you do not have a good contextual basis.”
While hospitalist groups typically try to measure physician workload in terms of the number of patient encounters or work RVUs, the second study makes a strong case for developing a much more nuanced model of hospitalist workload “and for figuring out which local characteristics within a hospital and program might be modified to allow doctors to safely see more patients.
Dr. Michtalik spoke with Today’s Hospitalist.
With so much attention paid to resident work hours and nurse-patient ratios, why hasn’t physician workload been studied before?
I think one factor is physician autonomy. While there are regulations in place for nursing and resident work hours, physicians function as independent practitioners once they complete training. Then there’s the issue of availability. Particularly in rural communities, there often aren’t enough physicians to provide continuous coverage if specific workhour regulations were applied to attending physicians.
There have been some studies looking at critical care physicians, with recommendations for ICU physician staffing in terms of safety and quality. And we know that increased experience can potentially improve outcomes for our surgical colleagues. But hospitalists might be on the opposite end of that spectrum, where increased volumes may lead to worse outcomes once a certain threshold is passed.
Were you surprised by the number of factors that hospitalists say affect their ability to handle more volume?
We were very impressed with the versatility of programs. For example, many programs use dedicated admitting teams to streamline patient care, but we found subcategories within that type of innovation.
Some institutions use dedicated admitting teams that are responsible for patients they admit that entire day. Other institutions use dedicated admitters only until 2 p.m. and then transition admissions to daytime rounders, so they have admitting responsibilities during only certain hours.
Of all the different workload variables that you identified, which are the most easily modified?
The most modifiable type is team structure. That would include, for example, adding midlevel providers or using geographic localization.
Some of the more difficult and resource-intensive options include nonshared caseworkers assigned to a particular service, electronic medical records and separate procedure services. One of our future goals is to examine the use and impact of each of these potential strategies.
What are you doing at Johns Hopkins to accommodate higher volumes?
One priority is geographical localization of patients within a unit or floor. We’re also exploring the use of separate admitting teams so that daytime physicians can take care of patients they’ve admitted, instead of being called back again throughout the day to the ED.
I also moonlight at several Hopkins’ facilities, and one of those hospitals does not have a separate admitting team. Instead, hospitalists alternate admissions, but if one physician is having a critical situation on the floor, he or she can have a subsequent admission assigned to a different doctor. That way, patient care is not interrupted in the ED, and doctors have the flexibility to respond to floor emergencies.
Did you identify best practices that hospitalist programs can use to streamline workload?
That’s where we envision this work heading. Our goal in subsequent studies is to determine the association between workload and objective outcomes that affect patient care and hospital finances. Some of those outcomes, such as patient satisfaction and readmission rates, are now part of value-based purchasing.
Overall, we want to determine possible interventions that could improve patient safety or care quality, or identify risk factors that could alert programs or institutions to a potential workload issue.
With so many different variables affecting workload, what should individual programs be doing?
One key aspect in all of this is front line provider input. No news does not necessarily mean good news. You have to actively seek out what is occurring in your program and hospital to see if there is a workload issue and how it affects patient care safety and quality. You can then take that information and develop strategies to both monitor workload and decide on interventions. The goal is to balance safety, quality and efficiency.
Phyllis Maguire is Executive Editor of Todays Hospitalist.
