Update From My Clinical Research Project Last Summer At Stanford Hospital

Specific Aims
  1. By improving the documentation of malnutrition in our hospital system through a new protocol, we aim to more accurately reflect the true state of malnutrition in our documentation and improve the observed to expected mortality ratio.
  2. In addition to assessing the state of malnutrition, we are interested in reducing the workload of our attending physicians by reducing the number of indeterminate malnutrition diagnosis
  1. Additionally, we want to assess the comfort and confidence of our interns, residents, and attendings in detecting and diagnosing malnutrition as well as their utilization of the group note, a team-based record (our intervention) for the patient note.

Preliminary Results

The key in analyzing the success of our study’s intervention relies on the query rate. The query rate represents the diagnoses that were not made or completed prior to the patient’s discharge and subsequent billing and reimbursement. When the patient is discharged, the Clinical Documentation Improvement team (CDI) looks over the note, and based on the available information will make a query to the attending physician on a patient if they feel a diagnosis is missed or has not been made. The query system is inherently a time sink to providers. Given its obvious coding, billing and reimbursement importance, answering queries is a priority at Stanford Hospital. There is a 100% response rate, but answering queries is a laborious clerical task that competes with physician’s clinical time. Once a query is made, additional time must be taken to go back to the electronic health record to review the patient's chart and notify the CDI team that the query was answered.  An average of 15 to 20 minutes is reported to be spent addressing one query.

However, these queries have clear upsides: they can improve patient record integrity, continuity of care, patient safety, quality measure capture and population health data capture. In this study our goal is for the group note to improve the documentation of malnutrition. Malnutrition was frequently the number one queried item in the internal medicine division of the hospital. Malnutrition was queried an average of 18-20 times per month over the past three years, which is quite high for internal medicine.

After implementing the group note intervention in July, 2019, we have observed a distinct drop in the query rate for malnutrition (Figure 1). There has been a 55% drop in the number of malnutrition queries for August (8) and September (9) compared to the numbers from last three (3) previous months: July (18), June (18), May (17). We are awaiting on results from October onward and plan to analyze the data from the coming months to identify whether this change was spurious or if the trend will continue.

Figure 1. Medicine Top Queries from July 2019 to September 2019

Although the raw numbers look promising, our future analysis will require observing query rate as a percentage of the patients in the division to understand the true implications of the effect of the intervention.

In surveying the doctors on internal medicine, it was quite clear that our intervention had good penetration and that more than 92% of those surveyed (which represent more than 60% of the new intern class at Stanford hospital) reported using the new progress note template (our intervention) (Table 1). Although these doctors reported using the progress note template there was a significant amount of discomfort in diagnosing protein calorie malnutrition and distinguishing between the categories of mild, moderate and severe PCM (Table 2). More than 86% of response indicated average or below average comfort in distinguishing between the various categories.
Table 1. Are you currently using the progress note template? (n=38)
  Frequency Percent
No 3 7.9
Yes 35 92.1
Table 2. I am comfortable with diagnosing protein calorie malnutrition (PCM), including distinguishing among mild, moderate, and sever PCM categories. (n=38)
Comfort (5=most comfortable) Frequency Percent Cumulative Frequency Cumulative Percent
1 8 21.1 8 21.1
2 15 39.5 23 60.6
3 10 26.3 33 86.8
4 4 10.5 37 97.3
5 1 2.6 38 100.0

Significance of Results
Although these results are only preliminary, it is promising to see the initial reduction in the query rate. Our intervention allows both the registered dieticians (RDs) and the physicians to contribute to the group note. This is very important because based on our survey, physicians are not inherently comfortable making the correct malnutrition diagnosis for a patient. Looking at this, someone may think the clear solution would be to further educate physicians to improve their capability. However, in practice this could prove more difficult. Registered dieticians have been trained to make this diagnosis and they have frequently cited a number of concerns that complicate their ability to make the assessment. In the case of a physician, given their busy and compact training schedules adding an additional element though possible might prove challenging.

After looking at the survey data, it is clear that physicians heavily rely on dieticians to make the diagnosis, with over 97% of survey participants trusting RDs to make the diagnosis (Table 3).  Instead of creating more work for the busy doctors, incorporating more of the healthcare team through the wide-spread adoption of the group note has shown promise. If we can continue to streamline our documentation process, we can reduce the query rate, reducing physician workload and burnout. This method is easily replicable within other divisions of the hospital and can quickly become a scalable solution not just for malnutrition in other divisions but also for other queries like sepsis, pulmonary edema, and encephalopathy.  
Table 3. if a dietitian makes a diagnosis related to PCM, I trust his/her diagnosis. (n=38)
Trust (5=most trustful) Frequency Percent Cumulative Frequency Cumulative Percent
1 1 2.6 1 2.6
4 9 23.7 10 26.3
5 28 73.7 38 100.0

Plans for Future Work
Going forward from our preliminary results we will follow the query rate in medicine for the next six months. Analyzing the data as a percentage of the population on medicine will ground the data and show the intervention’s impact. Additionally, I want to see if the introduction of the group-note changes comfort level of the physicians making the malnutrition diagnosis as well as if the adoption of the new group template continues to remain high. A future direction of this work would be applying this group note template to other departments of the hospital. At Stanford, surgery frequently complains of high query rates and we would be interested to see if we can run a pilot for their division as well.

Posted by Yasaswi Vengalasetti on Jan 23, 2020 11:22 AM America/New_York