Enhancing Resident Note Documentation: A Quality Improvement Initiative to Accurately Capture Patient Complexity

Highlights

• •Peer reviewed notes increased CC/MCC capture and improved length of stay index.
• •Note templates provided a flexible model to improve documentation quality.
• •Physician/CDS teams showed improvement in capture of coding complexity.

Abstract

Provider notes serve as a critical component of physician workflow, documenting essential aspects of patient care while also fulfilling regulatory and billing requirements. With increasing documentation complexity introduced by the Centers for Medicare and Medicaid Services and the 2021 mandate for open access to clinical notes, physicians in training must develop skills to accurately document patient complexity. This quality improvement initiative aimed to enhance inpatient note documentation by internal medicine residents, focusing on improving the capture of medical complexity in coding and billing standards.

Our intervention included the development and implementation of a standardized progress note template, a structured scoring rubric, multidisciplinary rounds and curriculum integrating faculty and peer-led feedback. The study measured documentation improvements through rubric scores, Length of Stay Index (LOSi), and complications or comorbidities (CC) and major complications or comorbidities (MCC) capture rates.

Results demonstrated improvements in LOSi and enhanced CC/MCC capture, leading to improved institutional performance metrics. This initiative highlights the necessity of integrating formal note-writing training within residency curricula to meet evolving documentation demands.

Introduction

Provider notes are fundamental to patient care, serving as a portal for communication among a diverse audience of healthcare professionals. Historically, clinical notes were primarily used to serve as a teaching tool, consolidate patient information, and synthesize medical data.1, 2, 3 In 1995, the complexity of note writing skyrocketed with the introduction of clinical note documentation requirements by the Center for Medicare and Medicaid Services (CMS)^4,5 mandating specific care elements and language in medical records for reimbursement and compliance purposes. This complexity was further compounded in 2021, when legislation mandated open access to clinical notes through electronic health records, providing patients a direct portal to their medical information.⁶

These legislative changes have changed the landscape of provider notes, resulting in increased documentation burden for providers whose notes must serve a variety of audiences, including coding and billing analysts for CMS reporting. ^7,8 Given the high stakes of medical note documentation, it is important to consider how to teach effective note writing to physicians in training to appropriately capture patients’ severity of illness and complexity of care.

Medical school curriculums have traditionally introduced students to the written note as a tool for clinical reasoning.⁹ The undergraduate medical education (UME) curriculum focuses on teaching how to synthesize a patient’s initial presentation and diagnosis in a “History and Physical exam” (H&P) note and subsequent status updates with daily progress notes. Historically, at our institution, upper-level residents and/or attendings would review and provide feedback on intern progress notes. However, the hectic pace of inpatient care and the demands of the medical system have precluded this feedback system over the past decade. While internal medicine residents spend hours writing notes in the electronic health record ^10,11 there is inadequate formal education and feedback in our residency curricula.¹² Moreover, now that inpatient notes must conform to industry standards, an additional layer of knowledge and teaching must be added to equip residents for effective inpatient note writing.

Different strategies have been used to design and deliver education about note writing and some are successful at improving qualitative measures of progress notes.13, 14, 15 Previous efforts at our institution relied on individual faculty members knowledgeable in inpatient documentation rules to deliver traditional didactics to internal medicine residents. However, this failed to create sustainable improvements in the appropriate capture of patient complexity, negatively impacting reimbursement and institutional scoring outcomes. Thus, we initiated a quality improvement project to improve inpatient note documentation by interns, with the goal of accurately capturing true patient complexity in coding and billing standards.

Methods

Project Design:

In devising this quality improvement project, we aimed to improve the capture of patient complexity in intern’s documented inpatient daily progress notes on our inpatient medicine services. Based on a previously validated tool reported in the literature of optimal notes, ¹⁶ a rubric ¹⁷ and a standardized note template were created. We specifically aimed to: (1) Formally train residents and interns on effective note writing (2) Evaluate interns’ daily progress notes scored by a rubric and (3) Measure the impact of the documented notes on our medical system as a whole through length of stay (LOS) metrics, Medicare Severity-Diagnosis Related Group (MS-DRG) weights and complication and morbidity data, and complications or comorbidities (CC) and major complications or comorbidities (MCC) capture.

In regard to LOS metrics, Vizient, Inc. is a non-profit healthcare services company that provides health systems with analytics and data that can be used to benchmark an institution’s performance compared to peers and inform performance improvement efforts. Vizient reports data related to LOS for academic medical centers. Each hospitalization is assigned an expected length of stay in days based on the MS-DRG calculated from written documentation in the medical record.⁶ Hospitals receive a fixed payment based on the average cost of care for patients in each MS-DRG. The expected LOS to observed LOS ratio, called LOS index (LOSi), can be used to compare performance across academic medical centers. These comparisons can contribute to an institution's reputation, reimbursement, and how it is viewed by prospective patients, trainees, and peer institutions. The information used to generate MS-DRGs and expected LOS is drawn by coders from the documented notes of physicians directly caring for patients. Thus, note completeness and accuracy are essential to appropriately reflect these measures.

Additionally, accurately capturing CC’s and MCC’s of patients directly impacts several hospital-associated factors, including MS-DRG level, Case Mix Index, and Mortality Index. Appropriately representing the patient’s overall state in these technical coding terms, garnered from inpatient notes, is essential as many of the quality and safety measures that impact hospital reputation and payments are risk stratified. When the CC/MCC capture rate is lower than it should be due to inadequate documentation of complex patient problems, the mortality rate index, among other factors, is skewed, appearing inappropriately high.

We hypothesized that standardizing formal teaching and note templates, alongside daily multidisciplinary collaboration, would improve documentation of internal medicine residents’ daily progress notes.¹⁸ This improvement would in turn increase the capture of medical complexity and high levels of clinical decision-making, thereby impacting LOSi and other performance metrics. 19, 20, 21, 22, 23, 24 Prior to this project, residents and interns used a variety of inpatient daily note templates within our electronic health record for documenting patient’s daily progress notes. These notes tended to be very lengthy, with redundant information, and often used as prompts for presentation purposes.

We postulated a custom, standardized note template alongside a scoring rubric would provide the necessary scaffolding for concise notes that address required note elements such as complexity of care and services rendered.25, 26, 27 One benefit of a standardized note is that it permits other necessary audiences, such as coders, to have a consistent view of inpatient notes, allowing quick retrieval of critical information. A second benefit of our templated note is that a “problem list” is built in and automatically updated daily. Therefore, when a provider adds diagnoses to the problem list in the EHR, the provider is prompted to (1) add appropriate specificity to the problem and (2) increase use of regulatory terminology requisite in coding. Finally, it was felt that spaced, standardized and interactive teaching and feedback on intern’s documented notes would improve consistency and use of the note template and rubric.^22,23

Study Intervention:

The University of North Carolina (UNC) Hospitals is a 1,001-bed public academic medical center in Chapel Hill, North Carolina. The UNC Department of Medicine is the largest department at the University of North Carolina, with 465 faculty members, over 200 residents and fellows, 12 divisions, and 11 research centers. The department is responsible for caring for hospitalized patients on 15 inpatient services.

We developed a prospective yearlong pilot with defined periods of observation and evaluation from July 2021 to June 2022 with an intent to impact the capture of medical diagnoses in clinical documentation and LOSi on 4 of our inpatient medicine teaching services. The study began with the development of a standardized daily progress note template embedded in our EHR for the interns to use for daily note writing (Figure 1). This template was created by a team of collaborators, including associate program directors, core faculty, inpatient medicine faculty, chief residents, and documentation and coding specialists. In conjunction with the rubric, the new note aimed to reduce redundancy and accurately reflect patient complexity and active diagnoses.

Figure 1.

Sample of note template with “@” or “{}” areas denoted for auto-populated entries.

Additionally, the standardized note made use of an existing EHR function called a problem list, which is auto populated at the top of every standardized progress note template (Figure 2). The problem list interface within the EHR prompts providers to clarify broad diagnoses. For example, if “anemia” was the diagnosis entered, the provider would then be prompted to specify the cause, if known, such as acute blood loss (CC), iron deficiency, hemolytic (CC), and so forth (Figure 3). This promotes accuracy and integrity of the documentation as well as easy retrieval for CDSs. The problem list was updated daily by interns to ensure active medical inpatient diagnoses were captured to properly reflect patient complexity, a component of the rubric score.

Figure 2.

Sample of auto-populated problem in the standardized daily progress note template.

Figure 3.

Problem list entry interface in HER.

First-year internal medicine residents, who wrote the majority of notes for inpatient medicine teams, participated in a curriculum run by core faculty members that included a workshop on note writing best practice principles and specifically how to use the interactive template and problem list for daily note writing. Third-year residents were trained as “peer educators” to score intern notes using a rubric (Table 1) that emphasized the use of the template, an updated problem list, minimizing copying information from prior days’ notes, use of succinct language, and minimizing copying forward information from prior days notes. They were also guided on how to provide productive feedback in a nonpunitive way.

Table 1.

Grading Rubric for Notes

UNC Progress Note Templating Rubric
Author of Note:	Date
Assessment:	YES (1)	NO (0)
Provides context of relevant PMH, including otherwise health Identifies principal/primary hospital problem Identifies current problems requiring ongoing active management Identifies progress of current primary problem(s) requiring hospitalization (better, same, worse) Avoids rehashing differential diagnosis once definitive diagnosis is established Reflects synthesis of symptoms and events without repetition of the subjective and objective portions of the note
Plan:	YES (1)	NO (0)
The plan is problem-oriented Problem List includes modifiers (CC’s and MCC’s i.e., IDDM with peripheral neuropathy) Omits irrelevant problems Identifies changes in management specific to day of note Provides rationale for new management changes Identifies new medications or changes to medications Identifies specific needs for discharge or transfer criteria
Subjective:	YES (1)	NO (0)
Interval Events are updated and accurate Includes 2-9 ROS
Objective:	YES (1)	NO (0)
Includes updated Vital Signs Physical Exam is Up to Date and includes 2 systems and 4 findings Labs/Studies are included only if relevant/pertinent
Overall:	YES (1)	NO (0)
Assessment/Plan is appropriate length for complexity of patient and does not include unnecessary information Writes dates, not days of the week Avoids obvious errors due to copy and paste No modifications or edits of the standard note template
TOTAL: (max 22)

Study#: 22-3298 was found to be exempt by the UNC Institutional Review Board.

Four inpatient medicine services, which were previously chosen by institutional leadership as part of a larger annual hospital organizational goal, piloted the project. These services, staffed by internal medicine residents and interns alongside faculty attendings from nephrology, hematology/oncology, hospital medicine, and general medicine, provided us with the opportunity to examine a diverse range of medical services and conditions across various specialties.

Using spaced repetition, volunteer third-year residents on elective rotations actively reviewed and scored inpatient daily progress notes and discussed areas of opportunity with clinical documentation specialists (CDS). They then met with the interns to review areas for improvement in their notes based on rubric scores. Additionally, throughout the project, daily communication and bimonthly meetings between faculty physician service line leaders and the CDSs resulted in a collaborative team that allowed for connection and sharing of best practices.

Results

Clinical Documentation

Templated progress notes written by first year residents were scored by “peer educators” before and after the intervention. The sample taken before scoring was exactly 1 year prior in an effort to control for experience in note writing due to time of year. Overall there was modest improvement in documentation (Table 2) based on rubric scores.

Table 2.

Rubric Scores for Notes Written Before and After Curriculum Implementation

Team	Pre-intervention (n)	Pre-intervention mean rubric score	Post-intervention (n)	Post-intervention mean rubric score	p-value
A Nephrology	20	17.6	21	19.6	.007
B Heme/Onc	39	17.15	51	18.1	.0911
C Hospitalist	20	16.32	35	16.6	.7
D General Medicine	20	15.06	27	17.4	.04

*Maximum rubric score = 22.LOS indices (observed, expected, index).

The LOSi for the participating services during the study period decreased from a baseline of 1.11 to 1.00. While the actual observed LOS on the participating services increased from 7.70 to 8.09 days , the expected LOS increased to a greater proportion from 6.92 to 8.10, thus leading to an improved LOSi. The LOSi on all other services, excluding the pilot services, in the medical center also improved but showed a smaller decrease from 1.17 to 1.14. These values were inclusive of Vizient outliers so reflective of all patients discharged during the intervention.

Complication and comorbidity coding:

The percentage of discharges falling into each MS-DRG complication and comorbidity category was also calculated for the year prior to intervention and for the pilot year. There was also an increase in cases coded at the highest complexity category on the participating services. This increase ranged from 2.96% to 21.78%. The team with the smallest increase (Team B) participated in training and implemented best coding practices prior to this study and still benefited from this pilot. The other teams ranged from 2.96-21.78% increase in complexity of coding (Figure 4). Additionally, there was an increase across all teams in average MS-DRG weight compared to the year before (Table 3).

Figure 4.

% Increase in cases of comorbidity coding.

Table 3.

Average MS-DRG Weight Comparison Pre and Post-Intervention

Medical Service	FY2021	FY2022
Team A (Nephrology)
Avg DRG Weight	1.78	1.79
Distinct encounters	785	670
Team B (Hematology/Oncology)
Avg DRG Weight	2.17	2.25
Distinct encounters	1707	1832
Team C (Hospital medicine)
Avg DRG Weight	1.63	1.85
Distinct encounters	772	707
Team D (General Medicine)
Avg DRG Weight	1.71	1.89
Distinct encounters	739	670

Avg. DRG Weight and distinct count of Encounter Number in Numeric broken down by Discharge Date/Time Year vs. Hospital Service.

The data is filtered on Location Name, Code Set Name / DRG Scheme and Patient Class.

The Location Name filter keeps UNCH and UNCH REHAB.

The Code Set Name / DRG Scheme filter keeps MS-DRG V37 (FY2020), MS-DRG V38 (FY2021), MS-DRG V39 (FY2022) and MS-DRG V40 (FY2023). The Patient Class filter keeps INPATIENT, INPATIENT PSYCH and INPATIENT REHAB. The view is filtered on Hospital Service, which keeps 6 members.

Discussion

In our study, we aimed to enhance the alignment of interns’ written daily inpatient progress notes with requirements needed for accurate coding and billing capture. Our quality improvement project focused on trainees in a multidisciplinary, collaborative approach to medical note writing which positively influenced our hospital metrics. Overall, the results showed that there was improvement in coding complexity due to the multidisciplinary approach and peer feedback process. The influence of the intervention on improving LOSi metrics and CC/MCC capture is meaningful to the institution by contributing to reputation and reimbursement.

The followed hospital metrics in the study, including LOSi, CC/MCC capture and DRG weight, all improved to varying degrees. While Team B, heme/onc, had a lower increase in CC/MCC capture, this could have been due to the fact that this team had already, independently, implemented best practices targeted and better capturing complexity prior to the study rollout. Team A, nephrology, also had a more modest rise, which is conjectured to be due to the large portion of dialysis patients on the service which is an MCC in most cases. These 2 services also had the largest proportion of patients in the highest CC/MCC prior to the study and therefore likely had less room for improvement.

While utilizing upper-level residents is not necessarily new, this feedback system in note writing had become obsolete in our institution due to the high demands of inpatient care. However, unique to this project, the upper-level residents were electively supporting interns as a non-clinical rotation, therefore not actively delivering care to the patients on the service, alleviating them from competing clinical duties. The peer review method not only provided interns with direct structured feedback but also kept upper-level residents abreast and cognizant of best practices in note writing techniques, helping them to improve their own documentation practices. Additionally, upper-level residents desiring to be more involved in formal teaching were provided meaningful opportunities to participate in teaching efforts while simultaneously improving hospital system metrics.

The multidisciplinary approach played a large role in our improved metrics. The ability to form a team with physicians and CDSs allowed each group to use their expertise in a collaborative manner, further impacting expected LOS and the LOSi. The improvement in the expected LOS outpacing the observed LOS indicates that our documentation efforts effectively improved the capture of patient complexity in a way that our medical teams could actively influence. Previous efforts at collaboration between these physicians and clinical documentation specialists faltered due to a perceived lack of understanding by resident physicians of the role of CDSs, concern about compliance rules, and lack of integration into physician’s daily workflow.

The improved CC/MCC capture is also due to interns consistently updating the active problem list, a component in their teaching and rubric scoring. As previously stated, problem list entry prompts providers to specify broad diagnoses, enhancing documentation accuracy. An example of how this impacts complexity capture is with the common term “volume overload.” Volume overload is nonspecific and the intern is prompted to state the cause of volume overload when entering it into the problem list. The provider then can choose from a variety of causes such as Chronic CHF or an Acute CHF Exacerbation. Chronic CHF adds a CC to the encounter and Acute CHF Exacerbation adds an MCC (this is not denoted in the problem list interface). Creating a note template with an embedded problem list increased the specificity of medical diagnoses and provided learners with a standardized progress note structure which allowed coders and others reviewing the medical record a consistent, standard format to retrieve information.

Our project designated 1 core faculty member to own the planning and implementation of the curriculum, including the training of “peer educators” and updating the standard note templates. Having a single faculty member oversee this project allows other faculty to dedicate more time to other competing demands of inpatient services. The flexible curriculum can be modified to work for internal medicine residency programs of different sizes and with varying rotation schedules. As previously noted, documentation requirements change with time. Fortunately, the templated note and grading rubric are easily modifiable by faculty, and adjustments are instantaneous for all users of the standard template, allowing real-time changes in documented notes (as was the case in 2023, when CMS, again, revised documentation guidelines).

Our study had several limitations; one limitation was the fact that there was no control service alongside the study services, thereby limiting the identification of bias or confounders. Thus, if other departments, such as coding, changed their process at the same time as our project, this would not be appreciated in our results. Secondly, our interns and residents rotated through multiple services in the hospital. Therefore, once an intern received formal teaching and direct note feedback from 1 of the pilot services, these skills hopefully will spread to other services they rotated through impacting metrics on other services. Apart from rubric scoring, there was no separate survey of learners to assess for improved understanding of coding principles.

Additionally, there were multiple different residents acting as “peer educators” with only 1 formal onboarding session provided on documentation and rubric grading. With a lack of longitudinal feedback for residents serving as “peer educators” this could cause a waning in retention and increased scoring variability in rubric over time. Moreover, residents selected which services they wanted to engage with as “peer educators.” Since there was no randomization, this may have led to differences in the core services they worked with, potentially introducing feedback bias. Finally, attendings on service were not integral to the project, therefore not necessarily overseeing the feedback or recommendations provided to the interns. The level of engagement of different attendings and services could also have altered outcomes.

Given the successes outlined above, our residency program plans to continue the program annually and expand from daily progress notes to include admission and discharge notes. In the future, further areas to study would be additional metrics to determine if notes improved beyond the aforementioned benchmarks such as note documentation length, reduction of copy and paste forward, time to complete charting and robust intern feedback. Additionally we would like to broaden our scope and project design to study the influences of this project on healthcare quality, healthcare team communication, physician 'cognitive load', and physician efficiency.

Conclusion

The implementations of templated notes along with senior resident peer review and coaching appear to improve clinical notes written by medicine interns thereby improving LOSi scores and facilitating billing and reimbursements. Academic hospitals could consider implementing similar protocols to enhance medical documentation.

Funding

This project received no external financial or non-financial support.

CRediT authorship contribution statement

De-Vaughn Williams: Writing – review & editing, Writing – original draft, Validation, Project administration, Methodology, Investigation. Scott Keller: Writing – review & editing, Data curation, Conceptualization. Jennifer Mcentee: Writing – review & editing, Project administration, Methodology, Conceptualization. Escher Howard-Williams: Writing – review & editing, Writing – original draft, Resources. Cristin M. Colford: Writing – review & editing, Writing – original draft, Resources, Project administration, Methodology, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.