1 / 21

the application of adjusted clinical group predictive models to the english population for risk adjustment, funding allo

Authors . Prof Jonathan Weiner, Johns Hopkins UniversitDr Klaus Lemke, Johns Hopkins UniversityDr Karen Kinder, ACG InternationalStephen Sutch, ACG International. Overview. IntroductionApplying the US ACG system to the English population and dataMethodsData preparation/cross-walksRegression modelsResultsConclusions and Recommendations.

paul2
Télécharger la présentation

the application of adjusted clinical group predictive models to the english population for risk adjustment, funding allo

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

    1. The application of Adjusted Clinical Group Predictive models to the English Population for risk adjustment, funding allocation Steve Sutch, MAppSc, BSc. Senior Analyst, ACG International The Johns Hopkins University On JHU faculty and a member of ACG TeamOn JHU faculty and a member of ACG Team

    3. Overview Introduction Applying the US ACG system to the English population and data Methods Data preparation/cross-walks Regression models Results Conclusions and Recommendations

    4. Introduction The current budgetary allocation formula to primary care trusts is based on capitation adjusted for demographic, socio-economic and population morbidity utilising aggregate and survey statistics. This study tested the feasibility of applying person based morbidity data as an alternative through the application of the ACG predictive models utilising hospitals secondary care data, primary care data, pharmacy and individual resident population data up to 50 million individuals.

    5. The risk measurement pyramid

    6. Objective To show the feasibility of adapting a US risk assessment system to UK inpatient and GP EMR data Morbidity and Pharmacy data UK uses ICD10, Read/CTV3 codes, BNF/DMD US uses IC9CM, NDC codes adaptation of US system via cross-walks “translation” of Read/CTV3 to ICD-10/ATC codes)

    7. Overview of Johns Hopkins ACG System The ACG - “Adjusted Clinical Group” system provides conceptually simple, statistically valid, and clinically relevant measures of need/risk for health services. The grouping process has been computerized. The “grouper” and “predictive modeling” software requires diagnosis information from encounter data, electronic medical records, or insurance records. ACGs are generally applied using all diagnoses describing the person. They do not focus on individual visits. Ideally they are derived from primary and specialty ambulatory contacts as well as inpatient There is a comprehensive ACG “suite” of risk/case-mix measures.

    8. The Johns Hopkins ACG System: An Expanding Suite of Measures Diagnosis (ICD) based: ADGs classify diagnoses into a limited number of clinically meaningful, but not disease-specific, morbidity groups. (For example “chronic unstable”) EDCs classify diagnoses based on specific diseases. They represent disease markers and can be used to determine disease prevalence (For example, Type I Diabetes, w/o complications). ACGs (Adjusted Clinical Groups) represent a single, mutually exclusive actuarial cell based on overall disease burden. (For example, 6-9 ADG Combinations, Age >34, 2 major ADGs) Can be combined to any number of Resource Utilization Bands or (RUBs) Pharmacy based: Rx-MG pharmacy risk markers, Rx-Morbidity Groups

    9. Method (1) The data sources were linked to form a single health record for each individual, comprising of inpatient, outpatient, Accident and Emergency (Emergency Room) and primary care activity data (including pharmacy) and related costs for two years. Crosswalks were designed to covert the primary care morbidity codes (READ codes) into ICD10 and the pharmacy codes into ATC (WHO Anatomical Therapeutic Chemical classification).

    10. Method (2) The data was grouped using the ACG software with the resultant grouping variables used to form a number of predictive models. Restrictive models, using secondary care data only Full models, using secondary and primary care morbidity data Combined morbidity and pharmacy models Parsimonious models, utilising 10-80 markers (non-parsimonious 300+ markers) The models were compared using R-squared and Mean Absolute Prediction Error (MAPE) statistics.

    11. Data UK inpatient (HES) and PCT Databases Year 1 inpatient ICD-10 clinical findings for all NHS registrants Year 2 inpatient costs for all NHS registrants Year 1 GP EMR data (GP clinical findings and medication prescriptions) for all residents in two selected PCTs PCT data augmented by HES inpatient data Year 2 “full” costs (inpatient and pharmacy) for all persons residing within PCT boundaries

    12. ACG Data process Import Files Patient File Diagnosis data file Pharmacy data file Output file (s) ACG Analysis tables

    13. Cross Walks “Adaptation” of ACG system to UK Read/CTV3 codes READ2/3 to ICD10 (to ADG, EDC) READ clinical findings were mapped to ICD-10 codes BNF/READ Drugs to ATC (to RxMG) Drug prescriptions were mapped to WHO ATC codes Subjected to clinical review, unmapped codes

    14. ACG Output Fields ACG Code, ADG Codes / Vector EDC Codes, MEDC Codes Resource Utilization Band Pharmacy Cost Band National Un/rescaled Weight Local Concurrent weight Hospital Dominant Count (ADG), Major ADG Count Frailty flag(s) Asthma, Congestive Heart Failure, COPD, Chronic Renal Failure, Depression, Diabetes, Hyperlipidemia, Hypertension, Ischemic Heart Disease, Low Back Pain ACG Predictive Model fields Total Cost RI, Probability High Cost, Pharmacy Cost RI, Probability High Pharmacy Cost,

    15. Data distribution

    16. Results

    17. Results

    18. Results The results for the morbidity models showed that the significant improvement in prediction over age/sex adjusted models, with increased improvement for the combined model (including pharmacy). The addition of primary care findings did not add to the explanatory power, but no primary care costs (except pharmacy) were applied The results were also compared to US data and showed higher predictive results for the English data.

    19. Conclusion The ACG grouping system can be applied to UK morbidity and pharmacy data from both secondary and primary care activity, with the predictive modelling demonstrating the strength of utilising such data for future budgetary allocation. A full costed model with primary care costs as well as secondary costs should allow for the greater realisation of benefits and allow the examination of a “total health system” for true World Class Commissioning. This study provides an initial approach to the introduction of funding allocation for chronic care management programs.

    20. ACGs reliably accept UK data inputs Diagnosis based (ICD-9 and ICD-10) has been converted to Read / CTV3 system with high level of completeness. Has been tested with GPRD (1.2 M) and UK morbidity survey and GP electronic medical records with and data from 5 PCTs (over 600K patients) Has been tested with secondary data (Over 12 million HES) We have completed conversion of Rx-MGs (pharmacy risk measures) to WHO ATCs and Read/CTV-3

    21. Web Site / Contacts ACG Web Site: www.acg.jhsph.edu Contacts: Dr. Karen Kinder-Siemens, Director International ACG Operations (kkinder@jhsph.edu) Steve Sutch (UK ACG Sr. Analyst) (steve@sutchconsulting.com) Professor Jonathan Weiner (jweiner@jhsph.edu)

More Related