Optimizing the start time of statin therapy for patients with diabetes

Brian T. Denton, Murat Kurt, Nilay D Shah, Sandra C. Bryant, Steven A. Smith

Research output: Contribution to journalArticle

33 Citations (Scopus)

Abstract

Background. Clinicians often use validated risk models to guide treatment decisions for cardiovascular risk reduction. The most common risk models for predicting cardiovascular risk are the UKPDS, Framingham, and Archimedes models. In this article, the authors propose a model to optimize the selection of patients for statin therapy of hypercholesterolemia, for patients with type 2 diabetes, using each of the risk models. For each model, they evaluate the role of age, gender, and metabolic state on the optimal start time for statins. Method. Using clinical data from the Mayo Clinic electronic medical record, the authors construct a Markov decision process model with health states composed of cardiovascular events and metabolic factors such as total cholesterol and high-density lipoproteins. They use it to evaluate the optimal start time of statin treatment for different combinations of cardiovascular risk models and patient attributes. Results. The authors find that treatment decisions depend on the cardiovascular risk model used and the age, gender, and metabolic state of the patient. Using the UKPDS risk model to estimate the probability of coronary heart disease and stroke events, they find that all white male patients should eventually start statin therapy; however, using Framingham and Archimedes models in place of UKPDS, they find that for male patients at lower risk, it is never optimal to initiate statins. For white female patients, the authors also find some patients for whom it is never optimal to initiate statins. Assuming that age 40 is the earliest possible start time, the authors find that the earliest optimal start times for UKPDS, Framingham, and Archimedes are 50, 46, and 40, respectively, for women. For men, the earliest optimal start times are 40, 40, and 40, respectively. Conclusions. In addition to age, gender, and metabolic state, the choice of cardiovascular risk model influences the apparent optimal time for starting statins in patients with diabetes.

Original languageEnglish (US)
Pages (from-to)351-367
Number of pages17
JournalMedical Decision Making
Volume29
Issue number3
DOIs
StatePublished - May 2009

Fingerprint

Hydroxymethylglutaryl-CoA Reductase Inhibitors
Cardiovascular Models
Therapeutics
Markov Chains
Electronic Health Records
Risk Reduction Behavior
Hypercholesterolemia
Type 2 Diabetes Mellitus
Patient Selection
HDL Cholesterol
Coronary Disease
Stroke

Keywords

  • Diabetes
  • Markov decision process
  • Statins

ASJC Scopus subject areas

  • Health Policy

Cite this

Optimizing the start time of statin therapy for patients with diabetes. / Denton, Brian T.; Kurt, Murat; Shah, Nilay D; Bryant, Sandra C.; Smith, Steven A.

In: Medical Decision Making, Vol. 29, No. 3, 05.2009, p. 351-367.

Research output: Contribution to journalArticle

Denton, Brian T. ; Kurt, Murat ; Shah, Nilay D ; Bryant, Sandra C. ; Smith, Steven A. / Optimizing the start time of statin therapy for patients with diabetes. In: Medical Decision Making. 2009 ; Vol. 29, No. 3. pp. 351-367.
@article{603205e656e3413d840d71ae6faf0a40,
title = "Optimizing the start time of statin therapy for patients with diabetes",
abstract = "Background. Clinicians often use validated risk models to guide treatment decisions for cardiovascular risk reduction. The most common risk models for predicting cardiovascular risk are the UKPDS, Framingham, and Archimedes models. In this article, the authors propose a model to optimize the selection of patients for statin therapy of hypercholesterolemia, for patients with type 2 diabetes, using each of the risk models. For each model, they evaluate the role of age, gender, and metabolic state on the optimal start time for statins. Method. Using clinical data from the Mayo Clinic electronic medical record, the authors construct a Markov decision process model with health states composed of cardiovascular events and metabolic factors such as total cholesterol and high-density lipoproteins. They use it to evaluate the optimal start time of statin treatment for different combinations of cardiovascular risk models and patient attributes. Results. The authors find that treatment decisions depend on the cardiovascular risk model used and the age, gender, and metabolic state of the patient. Using the UKPDS risk model to estimate the probability of coronary heart disease and stroke events, they find that all white male patients should eventually start statin therapy; however, using Framingham and Archimedes models in place of UKPDS, they find that for male patients at lower risk, it is never optimal to initiate statins. For white female patients, the authors also find some patients for whom it is never optimal to initiate statins. Assuming that age 40 is the earliest possible start time, the authors find that the earliest optimal start times for UKPDS, Framingham, and Archimedes are 50, 46, and 40, respectively, for women. For men, the earliest optimal start times are 40, 40, and 40, respectively. Conclusions. In addition to age, gender, and metabolic state, the choice of cardiovascular risk model influences the apparent optimal time for starting statins in patients with diabetes.",
keywords = "Diabetes, Markov decision process, Statins",
author = "Denton, {Brian T.} and Murat Kurt and Shah, {Nilay D} and Bryant, {Sandra C.} and Smith, {Steven A.}",
year = "2009",
month = "5",
doi = "10.1177/0272989X08329462",
language = "English (US)",
volume = "29",
pages = "351--367",
journal = "Medical Decision Making",
issn = "0272-989X",
publisher = "SAGE Publications Inc.",
number = "3",

}

TY - JOUR

T1 - Optimizing the start time of statin therapy for patients with diabetes

AU - Denton, Brian T.

AU - Kurt, Murat

AU - Shah, Nilay D

AU - Bryant, Sandra C.

AU - Smith, Steven A.

PY - 2009/5

Y1 - 2009/5

N2 - Background. Clinicians often use validated risk models to guide treatment decisions for cardiovascular risk reduction. The most common risk models for predicting cardiovascular risk are the UKPDS, Framingham, and Archimedes models. In this article, the authors propose a model to optimize the selection of patients for statin therapy of hypercholesterolemia, for patients with type 2 diabetes, using each of the risk models. For each model, they evaluate the role of age, gender, and metabolic state on the optimal start time for statins. Method. Using clinical data from the Mayo Clinic electronic medical record, the authors construct a Markov decision process model with health states composed of cardiovascular events and metabolic factors such as total cholesterol and high-density lipoproteins. They use it to evaluate the optimal start time of statin treatment for different combinations of cardiovascular risk models and patient attributes. Results. The authors find that treatment decisions depend on the cardiovascular risk model used and the age, gender, and metabolic state of the patient. Using the UKPDS risk model to estimate the probability of coronary heart disease and stroke events, they find that all white male patients should eventually start statin therapy; however, using Framingham and Archimedes models in place of UKPDS, they find that for male patients at lower risk, it is never optimal to initiate statins. For white female patients, the authors also find some patients for whom it is never optimal to initiate statins. Assuming that age 40 is the earliest possible start time, the authors find that the earliest optimal start times for UKPDS, Framingham, and Archimedes are 50, 46, and 40, respectively, for women. For men, the earliest optimal start times are 40, 40, and 40, respectively. Conclusions. In addition to age, gender, and metabolic state, the choice of cardiovascular risk model influences the apparent optimal time for starting statins in patients with diabetes.

AB - Background. Clinicians often use validated risk models to guide treatment decisions for cardiovascular risk reduction. The most common risk models for predicting cardiovascular risk are the UKPDS, Framingham, and Archimedes models. In this article, the authors propose a model to optimize the selection of patients for statin therapy of hypercholesterolemia, for patients with type 2 diabetes, using each of the risk models. For each model, they evaluate the role of age, gender, and metabolic state on the optimal start time for statins. Method. Using clinical data from the Mayo Clinic electronic medical record, the authors construct a Markov decision process model with health states composed of cardiovascular events and metabolic factors such as total cholesterol and high-density lipoproteins. They use it to evaluate the optimal start time of statin treatment for different combinations of cardiovascular risk models and patient attributes. Results. The authors find that treatment decisions depend on the cardiovascular risk model used and the age, gender, and metabolic state of the patient. Using the UKPDS risk model to estimate the probability of coronary heart disease and stroke events, they find that all white male patients should eventually start statin therapy; however, using Framingham and Archimedes models in place of UKPDS, they find that for male patients at lower risk, it is never optimal to initiate statins. For white female patients, the authors also find some patients for whom it is never optimal to initiate statins. Assuming that age 40 is the earliest possible start time, the authors find that the earliest optimal start times for UKPDS, Framingham, and Archimedes are 50, 46, and 40, respectively, for women. For men, the earliest optimal start times are 40, 40, and 40, respectively. Conclusions. In addition to age, gender, and metabolic state, the choice of cardiovascular risk model influences the apparent optimal time for starting statins in patients with diabetes.

KW - Diabetes

KW - Markov decision process

KW - Statins

UR - http://www.scopus.com/inward/record.url?scp=67649240453&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67649240453&partnerID=8YFLogxK

U2 - 10.1177/0272989X08329462

DO - 10.1177/0272989X08329462

M3 - Article

C2 - 19429836

AN - SCOPUS:67649240453

VL - 29

SP - 351

EP - 367

JO - Medical Decision Making

JF - Medical Decision Making

SN - 0272-989X

IS - 3

ER -