Related Experiment Videos
A patient-based system for describing ambulatory medicine practices using diagnosis clusters
B C Williams1, J T Philbrick, D M Becker
1Department of Internal Medicine, University of Virginia Health Sciences Center, Charlottesville.
This study aimed to create a classification system for describing the health conditions of patients seen in ambulatory medicine practices. The researchers developed 100 diagnosis clusters based on patient problem lists from a university practice and tested the system in a community practice. They reviewed a 5% sample of patient records to check the accuracy of the problem lists. The system successfully categorized 82% of university and 87% of community patient problems. The average number of health issues per patient was 6.1 in the university and 4.4 in the community. Common clusters included hypertension, obesity, and diabetes. The system captured most chronic conditions, with only 18% of university patients missing one or more important problems. The authors suggest this system can help in medical education, research, and policy-making by providing a structured way to analyze patient health data.
Area of Science:
- Ambulatory medicine practice analysis
- Medical education and clinical research
- Health services research
Background:
Understanding the clinical content of ambulatory medicine practices has been a challenge due to the variability in patient conditions and documentation methods. Prior research has shown that problem lists can help summarize patient health status, but no standardized system existed for categorizing these problems. That uncertainty drove the development of a classification system that could be applied across different practice settings. No prior work had resolved how to translate problem lists into a structured format for analysis. This gap motivated the creation of a system based on diagnosis clusters. Existing tools were limited in their ability to capture the full breadth of patient health issues. Researchers needed a method that could be used in both academic and community settings. The lack of a consistent framework made it difficult to compare practices or assess patient care quality. This study aimed to address those limitations by developing a patient-based classification system.
Purpose Of The Study:
The aim of this study was to create a classification system that could describe the clinical content of ambulatory medicine practices using diagnosis clusters. The researchers wanted to determine whether such a system could be applied consistently across different practice types. They focused on problem lists from patient records as the primary data source. The motivation was to improve the accuracy and utility of these lists for research and education. The study sought to evaluate the feasibility of using diagnosis clusters to summarize patient health. The researchers also aimed to assess the completeness of problem lists in capturing chronic conditions. They wanted to ensure the system could be used in both university and community settings. This approach could support better data analysis and inform public health strategies.
Main Methods:
The researchers developed a system of 100 diagnosis clusters by reviewing computerized problem lists from a university practice. They then applied this system to problem lists from a community practice. A 5% random sample of patients from the university practice was selected for chart review. This step was used to evaluate the accuracy of the computerized problem lists. The study included all patients seen in the university practice and those seen two or more times in the community practice. The time frame was a single year for both practices. The researchers assigned problems to diagnosis clusters and calculated the percentage of problems that could be categorized. They compared the number of problems and clusters per patient between the two practice settings.
Main Results:
Out of 27,634 problems listed for university patients, 22,629 (82%) were assigned to diagnosis clusters. For community patients, 4,924 out of 5,648 problems (87%) were assigned. The mean number of problems per patient was 6.1 in the university practice and 4.4 in the community practice. The mean number of diagnosis clusters per patient was 4.5 and 3.6, respectively. HYPERTENSION, SYMPTOM OR SIGN, OBESITY, and DIABETES were among the ten most common clusters in both practices. Only 18% of patient problem lists in the university practice omitted one or more chronic conditions. The system effectively described the clinical content of both practice types. The results suggest the system can be used for medical education and research purposes.
Conclusions:
The authors suggest that the diagnosis cluster system effectively described the clinical content of two types of internal medicine practices. They propose that this system has applications in medical education, epidemiology, and health services research. The results indicate that the system can be used to summarize patient health data efficiently. The authors suggest that the system supports the analysis of ambulatory medicine practices. They propose that the system can help identify common health issues across different patient populations. The findings suggest that the system can improve the accuracy of problem lists. The authors suggest that the system can be used to inform public policy decisions. They propose that the system can support better data collection and analysis in clinical settings.
Frequently Asked Questions
The system assigned 82% of university and 87% of community patient problems to diagnosis clusters.
The clusters were created by reviewing computerized problem lists from a university practice.
To assess the accuracy of the computerized problem lists in capturing chronic conditions.
They summarize the clinical content of patient records for analysis and comparison.
4.5 diagnosis clusters per patient in the university practice.
The authors suggest it can be used for medical education, epidemiology, and public policy.