Related Experiment Video
Updated: Apr 11, 2026
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Evaluating a Model to Predict Primary Care Physician-Defined Complexity in a Large Academic Primary Care
Clemens S Hong1, Steven J Atlas2, Jeffrey M Ashburner2
1General Medicine Division, Massachusetts General Hospital, 50 Staniford Street, 9th Floor, Boston, MA, 02114, USA. cshong@partners.org.
Background:
Improving the ability to risk-stratify patients is critical for efficiently allocating resources within healthcare systems.
Objective:
The purpose of this study was to evaluate a physician-defined complexity prediction model against outpatient Charlson score (OCS) and a commercial risk predictor (CRP).
Design:
Using a cohort in which primary care physicians reviewed 4302 of their adult patients, we developed a predictive model for estimated physician-defined complexity (ePDC) and categorized our population using ePDC, OCS and CRP.
Participants:
143,372 primary care patients in a practice-based research network participated in the study.
Main Measures:
For all patients categorized as complex in 2007 by one or more risk-stratification method, we calculated the percentage of total person time from 2008-2011 for which eligible cancer screening was incomplete, HbA1c was ≥ 9 %, and LDL was ≥ 130 mg/dl (in patients with cardiovascular disease). We also calculated the number of emergency department (ED) visits and hospital admissions per person year (ppy).
Key Results:
There was modest agreement among individuals classified as complex using ePDC compared with OCS (36.7 %) and CRP (39.6 %). Over 4 follow-up years, eligible ePDC-complex patients had higher proportions (p < 0.001) of time with: incomplete cervical (17.8 % vs. 13.3 % for OCS; 19.4 % vs. 11.2 % for CRP), breast (21.4 % vs. 14.9 % for OCS; 22.7 % vs. 15.0 % for CRP), and colon (25.9 % vs. 18.7 % for OCS; 27.0 % vs. 18.2 % for CRP) cancer screening; HbA1c ≥ 9 % (15.6 % vs. 8.1 % for OCS; 15.9 % vs. 6.9 % for CRP); and LDL ≥ 130 mg/dl (12.4 % vs. 7.9 % for OCS; 11.8 % vs 9.0 % for CRP). ePDC-complex patients had higher rates (p < 0.003) of: ED visits (0.21 vs. 0.11 ppy for OCS; 0.17 vs. 0.15 ppy for CRP), and admissions in patients 45-64 and ≥ 65 years old (0.11 vs. 0.10 ppy AND 0.24 vs. 0.21 ppy for OCS).
Conclusion:
Our measure for estimated physician-defined complexity compared favorably to commonly used risk-prediction approaches in identifying future suboptimal quality and utilization outcomes.
Related Concept Videos
01:15Methods of Documentation VI: Case Management Model
For example, a patient with a chronic...
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
00:59Model-Independent Approaches for Pharmacokinetic Data: Noncompartmental Analysis
One important characteristic of noncompartmental analyses is that drug exposure increases proportionally with increasing doses. This...
01:14Multicompartment Models: Overview
These models offer a more comprehensive representation of drug behavior in the body than one-compartment models. They accommodate the complexity of drug distribution,...
01:26Methods of Documentation II: POMR
01:14Model Approaches for Pharmacokinetic Data: Compartment Models
Two primary types of compartment models are recognized: mammillary and catenary. The more...