2015
Chhatwal, Jagpreet; Mathisen, Michael; Kantarjian, Hagop
Are high drug prices for hematologic malignancies justified? A critical analysis Journal Article
In: Cancer, vol. 121, no. 19, pp. 3372–3379, 2015, ISSN: 1097-0142.
@article{pmid26102457,
title = {Are high drug prices for hematologic malignancies justified? A critical analysis},
author = {Jagpreet Chhatwal and Michael Mathisen and Hagop Kantarjian},
doi = {10.1002/cncr.29512},
issn = {1097-0142},
year = {2015},
date = {2015-10-01},
journal = {Cancer},
volume = {121},
number = {19},
pages = {3372--3379},
abstract = {In the past 15 years, treatment outcomes for hematologic malignancies have improved substantially. However, drug prices have also increased drastically. This commentary examines the value of the treatment of hematologic malignancies at current prices in the United States through a reanalysis of a systematic review evaluating 29 studies of 9 treatments for 4 hematologic malignancies. Incremental cost-effectiveness ratios (ICERs) were calculated on the basis of drug prices in the United States in 2014. Sixty-three percent of the studies (15 of 24) had ICERs higher than $50,000 per quality-adjusted life-year (QALY), the benchmark widely used by health economists to define cost-effectiveness. In studies evaluating the current standard-of-care treatments for chronic myeloid leukemia, the ICERs for tyrosine kinase inhibitors versus hydroxyurea or interferon ranged from $210,000 to $426,000/QALY. The lower ICER values were mostly obtained from 11 studies evaluating rituximab, which was approved by the Food and Drug Administration in 1997 (ICER range, $37,000-$69,000/QALY). In conclusion, the costs of the majority of new treatments for hematologic cancers are too high to be deemed cost-effective in the United States.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In the past 15 years, treatment outcomes for hematologic malignancies have improved substantially. However, drug prices have also increased drastically. This commentary examines the value of the treatment of hematologic malignancies at current prices in the United States through a reanalysis of a systematic review evaluating 29 studies of 9 treatments for 4 hematologic malignancies. Incremental cost-effectiveness ratios (ICERs) were calculated on the basis of drug prices in the United States in 2014. Sixty-three percent of the studies (15 of 24) had ICERs higher than $50,000 per quality-adjusted life-year (QALY), the benchmark widely used by health economists to define cost-effectiveness. In studies evaluating the current standard-of-care treatments for chronic myeloid leukemia, the ICERs for tyrosine kinase inhibitors versus hydroxyurea or interferon ranged from $210,000 to $426,000/QALY. The lower ICER values were mostly obtained from 11 studies evaluating rituximab, which was approved by the Food and Drug Administration in 1997 (ICER range, $37,000-$69,000/QALY). In conclusion, the costs of the majority of new treatments for hematologic cancers are too high to be deemed cost-effective in the United States.
Seymour, Christopher W; Alotaik, Osama; Wallace, David J; Elhabashy, Ahmed E; Chhatwal, Jagpreet; Rea, Thomas D; Angus, Derek C; Nichol, Graham; Kahn, Jeremy M
County-Level Effects of Prehospital Regionalization of Critically Ill Patients: A Simulation Study Journal Article
In: Crit Care Med, vol. 43, no. 9, pp. 1807–1815, 2015, ISSN: 1530-0293.
@article{pmid26102251,
title = {County-Level Effects of Prehospital Regionalization of Critically Ill Patients: A Simulation Study},
author = {Christopher W Seymour and Osama Alotaik and David J Wallace and Ahmed E Elhabashy and Jagpreet Chhatwal and Thomas D Rea and Derek C Angus and Graham Nichol and Jeremy M Kahn},
doi = {10.1097/CCM.0000000000001133},
issn = {1530-0293},
year = {2015},
date = {2015-09-01},
journal = {Crit Care Med},
volume = {43},
number = {9},
pages = {1807--1815},
abstract = {OBJECTIVE: Regionalization may improve critical care delivery, yet stakeholders cite concerns about its feasibility. We sought to determine the operational effects of prehospital regionalization of nontrauma, nonarrest critical illness.nnSETTING: King County, Washington.nnDESIGN: Discrete event simulation study.nnPATIENTS: All 2006 hospital discharge data, linked to all adult, eligible patients transported by county emergency medical services agencies.nnINTERVENTIONS: We simulated active triage of high-risk patients to designated referral centers using a validated prehospital risk score; we studied three regionalization scenarios: 1) up triage, 2) up and down triage, and 3) up and down triage after reducing ICU beds by 25%. We determined the effect on patient routing, ICU occupancy at referral and nonreferral hospitals, and emergency medical services transport times.nnMEASUREMENTS AND MAIN RESULTS: A total of 119,117 patients were hospitalized at 11 nonreferral centers and 76,817 patients were hospitalized at three referral centers. Among 20,835 emergency medical services patients, 7,817 patients (43%) were eligible for up triage and 10,242 patients (57%) were eligible for down triage. At baseline, mean daily ICU bed occupancy was 61% referral and 47% at nonreferral hospitals. Up triage increased referral ICU occupancy to 68%, up and down triage to 64%, and up and down triage with bed reduction to 74%. Mean daily nonreferral ICU occupancy did not exceed 60%. Total emergency medical services transport time increased by less than 3% with up and down triage.nnCONCLUSIONS: Regionalization based on prehospital triage of the critically ill can allocate high-risk patients to referral hospitals without adversely affecting ICU occupancy or prehospital travel time.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
OBJECTIVE: Regionalization may improve critical care delivery, yet stakeholders cite concerns about its feasibility. We sought to determine the operational effects of prehospital regionalization of nontrauma, nonarrest critical illness.nnSETTING: King County, Washington.nnDESIGN: Discrete event simulation study.nnPATIENTS: All 2006 hospital discharge data, linked to all adult, eligible patients transported by county emergency medical services agencies.nnINTERVENTIONS: We simulated active triage of high-risk patients to designated referral centers using a validated prehospital risk score; we studied three regionalization scenarios: 1) up triage, 2) up and down triage, and 3) up and down triage after reducing ICU beds by 25%. We determined the effect on patient routing, ICU occupancy at referral and nonreferral hospitals, and emergency medical services transport times.nnMEASUREMENTS AND MAIN RESULTS: A total of 119,117 patients were hospitalized at 11 nonreferral centers and 76,817 patients were hospitalized at three referral centers. Among 20,835 emergency medical services patients, 7,817 patients (43%) were eligible for up triage and 10,242 patients (57%) were eligible for down triage. At baseline, mean daily ICU bed occupancy was 61% referral and 47% at nonreferral hospitals. Up triage increased referral ICU occupancy to 68%, up and down triage to 64%, and up and down triage with bed reduction to 74%. Mean daily nonreferral ICU occupancy did not exceed 60%. Total emergency medical services transport time increased by less than 3% with up and down triage.nnCONCLUSIONS: Regionalization based on prehospital triage of the critically ill can allocate high-risk patients to referral hospitals without adversely affecting ICU occupancy or prehospital travel time.
Elbasha, Elamin H; Chhatwal, Jagpreet
Characterizing Heterogeneity Bias in Cohort-Based Models Journal Article
In: Pharmacoeconomics, vol. 33, no. 8, pp. 857–865, 2015, ISSN: 1179-2027.
@article{pmid25851486,
title = {Characterizing Heterogeneity Bias in Cohort-Based Models},
author = {Elamin H Elbasha and Jagpreet Chhatwal},
doi = {10.1007/s40273-015-0273-z},
issn = {1179-2027},
year = {2015},
date = {2015-08-01},
journal = {Pharmacoeconomics},
volume = {33},
number = {8},
pages = {857--865},
abstract = {PURPOSE: Previous research using numerical methods suggested that use of a cohort-based model instead of an individual-based model can result in significant heterogeneity bias. However, the direction of the bias is not known a priori. We characterized mathematically the conditions that lead to upward or downward bias.nnMETHOD: We used a standard three-state disease progression model to evaluate the cost effectiveness of a hypothetical intervention. We solved the model analytically and derived expressions for life expectancy, discounted quality-adjusted life years (QALYs), discounted lifetime costs and incremental net monetary benefits (INMB). An outcome was calculated using the mean of the input under the cohort-based approach and the whole input distribution for all persons under the individual-based approach. We investigated the impact of heterogeneity on outcomes by varying one parameter at a time while keeping all others constant. We evaluated the curvature of outcome functions and used Jensen's inequality to determine the direction of the bias.nnRESULTS: Both life expectancy and QALYs were underestimated by the cohort-based approach. If there was heterogeneity only in disease progression, total costs were overestimated, whereas QALYs gained, incremental costs and INMB were under- or overestimated, depending on the progression rate. INMB was underestimated when only efficacy was heterogeneous. Both approaches yielded the same outcome when the heterogeneity was only in cost or utilities.nnCONCLUSION: A cohort-based approach that does not adjust for heterogeneity underestimates life expectancy and may underestimate or overestimate other outcomes. Characterizing the bias is useful for comparative assessment of models and informing decision making.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
PURPOSE: Previous research using numerical methods suggested that use of a cohort-based model instead of an individual-based model can result in significant heterogeneity bias. However, the direction of the bias is not known a priori. We characterized mathematically the conditions that lead to upward or downward bias.nnMETHOD: We used a standard three-state disease progression model to evaluate the cost effectiveness of a hypothetical intervention. We solved the model analytically and derived expressions for life expectancy, discounted quality-adjusted life years (QALYs), discounted lifetime costs and incremental net monetary benefits (INMB). An outcome was calculated using the mean of the input under the cohort-based approach and the whole input distribution for all persons under the individual-based approach. We investigated the impact of heterogeneity on outcomes by varying one parameter at a time while keeping all others constant. We evaluated the curvature of outcome functions and used Jensen's inequality to determine the direction of the bias.nnRESULTS: Both life expectancy and QALYs were underestimated by the cohort-based approach. If there was heterogeneity only in disease progression, total costs were overestimated, whereas QALYs gained, incremental costs and INMB were under- or overestimated, depending on the progression rate. INMB was underestimated when only efficacy was heterogeneous. Both approaches yielded the same outcome when the heterogeneity was only in cost or utilities.nnCONCLUSION: A cohort-based approach that does not adjust for heterogeneity underestimates life expectancy and may underestimate or overestimate other outcomes. Characterizing the bias is useful for comparative assessment of models and informing decision making.
Chhatwal, Jagpreet; He, Tianhua
Economic evaluations with agent-based modelling: an introduction Journal Article
In: Pharmacoeconomics, vol. 33, no. 5, pp. 423–433, 2015, ISSN: 1179-2027.
@article{pmid25609398,
title = {Economic evaluations with agent-based modelling: an introduction},
author = {Jagpreet Chhatwal and Tianhua He},
doi = {10.1007/s40273-015-0254-2},
issn = {1179-2027},
year = {2015},
date = {2015-05-01},
journal = {Pharmacoeconomics},
volume = {33},
number = {5},
pages = {423--433},
abstract = {Agent-based modelling (ABM) is a relatively new technique, which overcomes some of the limitations of other methods commonly used for economic evaluations. These limitations include linearity, homogeneity and stationarity. Agents in ABMs are autonomous entities, who interact with each other and with the environment. ABMs provide an inductive or 'bottom-up' approach, i.e. individual-level behaviours define system-level components. ABMs have a unique property to capture emergence phenomena that otherwise cannot be predicted by the combination of individual-level interactions. In this tutorial, we discuss the basic concepts and important features of ABMs. We present a case study of an application of a simple ABM to evaluate the cost effectiveness of screening of an infectious disease. We also provide our model, which was developed using an open-source software program, NetLogo. We discuss software, resources, challenges and future research opportunities of ABMs for economic evaluations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Agent-based modelling (ABM) is a relatively new technique, which overcomes some of the limitations of other methods commonly used for economic evaluations. These limitations include linearity, homogeneity and stationarity. Agents in ABMs are autonomous entities, who interact with each other and with the environment. ABMs provide an inductive or 'bottom-up' approach, i.e. individual-level behaviours define system-level components. ABMs have a unique property to capture emergence phenomena that otherwise cannot be predicted by the combination of individual-level interactions. In this tutorial, we discuss the basic concepts and important features of ABMs. We present a case study of an application of a simple ABM to evaluate the cost effectiveness of screening of an infectious disease. We also provide our model, which was developed using an open-source software program, NetLogo. We discuss software, resources, challenges and future research opportunities of ABMs for economic evaluations.
Chhatwal, Jagpreet; Kanwal, Fasiha; Roberts, Mark S; Dunn, Michael A
Cost-effectiveness and budget impact of hepatitis C virus treatment with sofosbuvir and ledipasvir in the United States Journal Article
In: Ann Intern Med, vol. 162, no. 6, pp. 397–406, 2015, ISSN: 1539-3704.
@article{pmid25775312,
title = {Cost-effectiveness and budget impact of hepatitis C virus treatment with sofosbuvir and ledipasvir in the United States},
author = {Jagpreet Chhatwal and Fasiha Kanwal and Mark S Roberts and Michael A Dunn},
doi = {10.7326/M14-1336},
issn = {1539-3704},
year = {2015},
date = {2015-03-01},
journal = {Ann Intern Med},
volume = {162},
number = {6},
pages = {397--406},
abstract = {BACKGROUND: Sofosbuvir and ledipasvir, which have recently been approved for treatment of chronic hepatitis C virus (HCV) infection, are more efficacious and safer than the old standard of care (oSOC) but are substantially more expensive. Whether and in which patients their improved efficacy justifies their increased cost is unclear.nnOBJECTIVE: To evaluate the cost-effectiveness and budget impact of sofosbuvir and ledipasvir.nnDESIGN: Microsimulation model of the natural history of HCV infection.nnDATA SOURCES: Published literature.nnTARGET POPULATION: Treatment-naive and treatment-experienced HCV population defined on the basis of HCV genotype, age, and fibrosis distribution in the United States.nnTIME HORIZON: Lifetime.nnPERSPECTIVE: Third-party payer.nnINTERVENTION: Simulation of sofosbuvir-ledipasvir compared with the oSOC (interferon-based therapies).nnOUTCOME MEASURES: Quality-adjusted life-years (QALYs), incremental cost-effectiveness ratios (ICERs), and 5-year spending on antiviral drugs.nnRESULTS OF BASE-CASE ANALYSIS: Sofosbuvir-based therapies added 0.56 QALY relative to the oSOC at an ICER of $55 400 per additional QALY. The ICERs ranged from $9700 to $284 300 per QALY depending on the patient's status with respect to treatment history, HCV genotype, and presence of cirrhosis. At a willingness-to-pay threshold of $100 000 per QALY, sofosbuvir-based therapies were cost-effective in 83% of treatment-naive and 81% of treatment-experienced patients. Compared with the oSOC, treating eligible HCV-infected persons in the United States with the new drugs would cost an additional $65 billion in the next 5 years, whereas the resulting cost offsets would be $16 billion.nnRESULTS OF SENSITIVITY ANALYSIS: Results were sensitive to drug price, drug efficacy, and quality of life after successful treatment.nnLIMITATION: Data on real-world effectiveness of new antivirals are lacking.nnCONCLUSION: Treatment of HCV is cost-effective in most patients, but additional resources and value-based patient prioritization are needed to manage patients with HCV.nnPRIMARY FUNDING SOURCE: National Institutes of Health.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: Sofosbuvir and ledipasvir, which have recently been approved for treatment of chronic hepatitis C virus (HCV) infection, are more efficacious and safer than the old standard of care (oSOC) but are substantially more expensive. Whether and in which patients their improved efficacy justifies their increased cost is unclear.nnOBJECTIVE: To evaluate the cost-effectiveness and budget impact of sofosbuvir and ledipasvir.nnDESIGN: Microsimulation model of the natural history of HCV infection.nnDATA SOURCES: Published literature.nnTARGET POPULATION: Treatment-naive and treatment-experienced HCV population defined on the basis of HCV genotype, age, and fibrosis distribution in the United States.nnTIME HORIZON: Lifetime.nnPERSPECTIVE: Third-party payer.nnINTERVENTION: Simulation of sofosbuvir-ledipasvir compared with the oSOC (interferon-based therapies).nnOUTCOME MEASURES: Quality-adjusted life-years (QALYs), incremental cost-effectiveness ratios (ICERs), and 5-year spending on antiviral drugs.nnRESULTS OF BASE-CASE ANALYSIS: Sofosbuvir-based therapies added 0.56 QALY relative to the oSOC at an ICER of $55 400 per additional QALY. The ICERs ranged from $9700 to $284 300 per QALY depending on the patient's status with respect to treatment history, HCV genotype, and presence of cirrhosis. At a willingness-to-pay threshold of $100 000 per QALY, sofosbuvir-based therapies were cost-effective in 83% of treatment-naive and 81% of treatment-experienced patients. Compared with the oSOC, treating eligible HCV-infected persons in the United States with the new drugs would cost an additional $65 billion in the next 5 years, whereas the resulting cost offsets would be $16 billion.nnRESULTS OF SENSITIVITY ANALYSIS: Results were sensitive to drug price, drug efficacy, and quality of life after successful treatment.nnLIMITATION: Data on real-world effectiveness of new antivirals are lacking.nnCONCLUSION: Treatment of HCV is cost-effective in most patients, but additional resources and value-based patient prioritization are needed to manage patients with HCV.nnPRIMARY FUNDING SOURCE: National Institutes of Health.
2014
Kabiri, Mina; Jazwinski, Alison B; Roberts, Mark S; Schaefer, Andrew J; Chhatwal, Jagpreet
The changing burden of hepatitis C virus infection in the United States: model-based predictions Journal Article
In: Ann Intern Med, vol. 161, no. 3, pp. 170–180, 2014, ISSN: 1539-3704.
@article{pmid25089861,
title = {The changing burden of hepatitis C virus infection in the United States: model-based predictions},
author = {Mina Kabiri and Alison B Jazwinski and Mark S Roberts and Andrew J Schaefer and Jagpreet Chhatwal},
doi = {10.7326/M14-0095},
issn = {1539-3704},
year = {2014},
date = {2014-08-01},
journal = {Ann Intern Med},
volume = {161},
number = {3},
pages = {170--180},
abstract = {BACKGROUND: Chronic hepatitis C virus (HCV) infection causes a substantial health and economic burden in the United States. With the availability of direct-acting antiviral agents, recently approved therapies and those under development, and 1-time birth-cohort screening, the burden of this disease is expected to decrease.nnOBJECTIVE: To predict the effect of new therapies and screening on chronic HCV infection and associated disease outcomes.nnDESIGN: Individual-level state-transition model.nnSETTING: Existing and anticipated therapies and screening for HCV infection in the United States.nnPATIENTS: Total HCV-infected population in the United States.nnMEASUREMENTS: The number of cases of chronic HCV infection and outcomes of advanced-stage HCV infection.nnRESULTS: The number of cases of chronic HCV infection decreased from 3.2 million in 2001 to 2.3 million in 2013. One-time birth-cohort screening beginning in 2013 is expected to identify 487,000 cases of HCV infection in the next 10 years. In contrast, 1-time universal screening could identify 933,700 cases. With the availability of highly effective therapies, HCV infection could become a rare disease in the next 22 years. Recently approved therapies for HCV infection and 1-time birth-cohort screening could prevent approximately 124,200 cases of decompensated cirrhosis, 78,800 cases of hepatocellular carcinoma, 126,500 liver-related deaths, and 9900 liver transplantations by 2050. Increasing the treatment capacity would further reduce the burden of HCV disease.nnLIMITATION: Institutionalized patients with HCV infection were excluded, and empirical data on the effectiveness of future therapies and on the future annual incidence and treatment capacity of HCV infection are lacking.nnCONCLUSION: New therapies for HCV infection and widespread implementation of screening and treatment will play an important role in reducing the burden of HCV disease. More aggressive screening recommendations are needed to identify a large pool of infected patients.nnPRIMARY FUNDING SOURCE: National Institutes of Health.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: Chronic hepatitis C virus (HCV) infection causes a substantial health and economic burden in the United States. With the availability of direct-acting antiviral agents, recently approved therapies and those under development, and 1-time birth-cohort screening, the burden of this disease is expected to decrease.nnOBJECTIVE: To predict the effect of new therapies and screening on chronic HCV infection and associated disease outcomes.nnDESIGN: Individual-level state-transition model.nnSETTING: Existing and anticipated therapies and screening for HCV infection in the United States.nnPATIENTS: Total HCV-infected population in the United States.nnMEASUREMENTS: The number of cases of chronic HCV infection and outcomes of advanced-stage HCV infection.nnRESULTS: The number of cases of chronic HCV infection decreased from 3.2 million in 2001 to 2.3 million in 2013. One-time birth-cohort screening beginning in 2013 is expected to identify 487,000 cases of HCV infection in the next 10 years. In contrast, 1-time universal screening could identify 933,700 cases. With the availability of highly effective therapies, HCV infection could become a rare disease in the next 22 years. Recently approved therapies for HCV infection and 1-time birth-cohort screening could prevent approximately 124,200 cases of decompensated cirrhosis, 78,800 cases of hepatocellular carcinoma, 126,500 liver-related deaths, and 9900 liver transplantations by 2050. Increasing the treatment capacity would further reduce the burden of HCV disease.nnLIMITATION: Institutionalized patients with HCV infection were excluded, and empirical data on the effectiveness of future therapies and on the future annual incidence and treatment capacity of HCV infection are lacking.nnCONCLUSION: New therapies for HCV infection and widespread implementation of screening and treatment will play an important role in reducing the burden of HCV disease. More aggressive screening recommendations are needed to identify a large pool of infected patients.nnPRIMARY FUNDING SOURCE: National Institutes of Health.
Ayvaci, Mehmet U S; Alagoz, Oguzhan; Chhatwal, Jagpreet; del Rio, Alejandro Munoz; Sickles, Edward A; Nassif, Houssam; Kerlikowske, Karla; Burnside, Elizabeth S
Predicting invasive breast cancer versus DCIS in different age groups Journal Article
In: BMC Cancer, vol. 14, pp. 584, 2014, ISSN: 1471-2407.
@article{pmid25112586,
title = {Predicting invasive breast cancer versus DCIS in different age groups},
author = {Mehmet U S Ayvaci and Oguzhan Alagoz and Jagpreet Chhatwal and Alejandro Munoz del Rio and Edward A Sickles and Houssam Nassif and Karla Kerlikowske and Elizabeth S Burnside},
doi = {10.1186/1471-2407-14-584},
issn = {1471-2407},
year = {2014},
date = {2014-08-01},
journal = {BMC Cancer},
volume = {14},
pages = {584},
abstract = {BACKGROUND: Increasing focus on potentially unnecessary diagnosis and treatment of certain breast cancers prompted our investigation of whether clinical and mammographic features predictive of invasive breast cancer versus ductal carcinoma in situ (DCIS) differ by age.nnMETHODS: We analyzed 1,475 malignant breast biopsies, 1,063 invasive and 412 DCIS, from 35,871 prospectively collected consecutive diagnostic mammograms interpreted at University of California, San Francisco between 1/6/1997 and 6/29/2007. We constructed three logistic regression models to predict the probability of invasive cancer versus DCIS for the following groups: women ≥ 65 (older group), women 50-64 (middle age group), and women < 50 (younger group). We identified significant predictors and measured the performance in all models using area under the receiver operating characteristic curve (AUC).nnRESULTS: The models for older and the middle age groups performed significantly better than the model for younger group (AUC = 0.848 vs, 0.778; p = 0.049 and AUC = 0.851 vs, 0.778; p = 0.022, respectively). Palpability and principal mammographic finding were significant predictors in distinguishing invasive from DCIS in all age groups. Family history of breast cancer, mass shape and mass margins were significant positive predictors of invasive cancer in the older group whereas calcification distribution was a negative predictor of invasive cancer (i.e. predicted DCIS). In the middle age group--mass margins, and in the younger group--mass size were positive predictors of invasive cancer.nnCONCLUSIONS: Clinical and mammographic finding features predict invasive breast cancer versus DCIS better in older women than younger women. Specific predictive variables differ based on age.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: Increasing focus on potentially unnecessary diagnosis and treatment of certain breast cancers prompted our investigation of whether clinical and mammographic features predictive of invasive breast cancer versus ductal carcinoma in situ (DCIS) differ by age.nnMETHODS: We analyzed 1,475 malignant breast biopsies, 1,063 invasive and 412 DCIS, from 35,871 prospectively collected consecutive diagnostic mammograms interpreted at University of California, San Francisco between 1/6/1997 and 6/29/2007. We constructed three logistic regression models to predict the probability of invasive cancer versus DCIS for the following groups: women ≥ 65 (older group), women 50-64 (middle age group), and women < 50 (younger group). We identified significant predictors and measured the performance in all models using area under the receiver operating characteristic curve (AUC).nnRESULTS: The models for older and the middle age groups performed significantly better than the model for younger group (AUC = 0.848 vs, 0.778; p = 0.049 and AUC = 0.851 vs, 0.778; p = 0.022, respectively). Palpability and principal mammographic finding were significant predictors in distinguishing invasive from DCIS in all age groups. Family history of breast cancer, mass shape and mass margins were significant positive predictors of invasive cancer in the older group whereas calcification distribution was a negative predictor of invasive cancer (i.e. predicted DCIS). In the middle age group--mass margins, and in the younger group--mass size were positive predictors of invasive cancer.nnCONCLUSIONS: Clinical and mammographic finding features predict invasive breast cancer versus DCIS better in older women than younger women. Specific predictive variables differ based on age.
Wallace, David J; Kahn, Jeremy M; Angus, Derek C; Martin-Gill, Christian; Callaway, Clifton W; Rea, Thomas D; Chhatwal, Jagpreet; Kurland, Kristen; Seymour, Christopher W
Accuracy of prehospital transport time estimation Journal Article
In: Acad Emerg Med, vol. 21, no. 1, pp. 9–16, 2014, ISSN: 1553-2712.
@article{pmid24552519,
title = {Accuracy of prehospital transport time estimation},
author = {David J Wallace and Jeremy M Kahn and Derek C Angus and Christian Martin-Gill and Clifton W Callaway and Thomas D Rea and Jagpreet Chhatwal and Kristen Kurland and Christopher W Seymour},
doi = {10.1111/acem.12289},
issn = {1553-2712},
year = {2014},
date = {2014-01-01},
journal = {Acad Emerg Med},
volume = {21},
number = {1},
pages = {9--16},
abstract = {OBJECTIVES: Estimates of prehospital transport times are an important part of emergency care system research and planning; however, the accuracy of these estimates is unknown. The authors examined the accuracy of three estimation methods against observed transport times in a large cohort of prehospital patient transports.nnMETHODS: This was a validation study using prehospital records in King County, Washington, and southwestern Pennsylvania from 2002 to 2006 and 2005 to 2011, respectively. Transport time estimates were generated using three methods: linear arc distance, Google Maps, and ArcGIS Network Analyst. Estimation error, defined as the absolute difference between observed and estimated transport time, was assessed, as well as the proportion of estimated times that were within specified error thresholds. Based on the primary results, a regression estimate was used that incorporated population density, time of day, and season to assess improved accuracy. Finally, hospital catchment areas were compared using each method with a fixed drive time.nnRESULTS: The authors analyzed 29,935 prehospital transports to 44 hospitals. The mean (± standard deviation [±SD]) absolute error was 4.8 (±7.3) minutes using linear arc, 3.5 (±5.4) minutes using Google Maps, and 4.4 (±5.7) minutes using ArcGIS. All pairwise comparisons were statistically significant (p < 0.01). Estimation accuracy was lower for each method among transports more than 20 minutes (mean [±SD] absolute error was 12.7 [±11.7] minutes for linear arc, 9.8 [±10.5] minutes for Google Maps, and 11.6 [±10.9] minutes for ArcGIS). Estimates were within 5 minutes of observed transport time for 79% of linear arc estimates, 86.6% of Google Maps estimates, and 81.3% of ArcGIS estimates. The regression-based approach did not substantially improve estimation. There were large differences in hospital catchment areas estimated by each method.nnCONCLUSIONS: Route-based transport time estimates demonstrate moderate accuracy. These methods can be valuable for informing a host of decisions related to the system organization and patient access to emergency medical care; however, they should be employed with sensitivity to their limitations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
OBJECTIVES: Estimates of prehospital transport times are an important part of emergency care system research and planning; however, the accuracy of these estimates is unknown. The authors examined the accuracy of three estimation methods against observed transport times in a large cohort of prehospital patient transports.nnMETHODS: This was a validation study using prehospital records in King County, Washington, and southwestern Pennsylvania from 2002 to 2006 and 2005 to 2011, respectively. Transport time estimates were generated using three methods: linear arc distance, Google Maps, and ArcGIS Network Analyst. Estimation error, defined as the absolute difference between observed and estimated transport time, was assessed, as well as the proportion of estimated times that were within specified error thresholds. Based on the primary results, a regression estimate was used that incorporated population density, time of day, and season to assess improved accuracy. Finally, hospital catchment areas were compared using each method with a fixed drive time.nnRESULTS: The authors analyzed 29,935 prehospital transports to 44 hospitals. The mean (± standard deviation [±SD]) absolute error was 4.8 (±7.3) minutes using linear arc, 3.5 (±5.4) minutes using Google Maps, and 4.4 (±5.7) minutes using ArcGIS. All pairwise comparisons were statistically significant (p < 0.01). Estimation accuracy was lower for each method among transports more than 20 minutes (mean [±SD] absolute error was 12.7 [±11.7] minutes for linear arc, 9.8 [±10.5] minutes for Google Maps, and 11.6 [±10.9] minutes for ArcGIS). Estimates were within 5 minutes of observed transport time for 79% of linear arc estimates, 86.6% of Google Maps estimates, and 81.3% of ArcGIS estimates. The regression-based approach did not substantially improve estimation. There were large differences in hospital catchment areas estimated by each method.nnCONCLUSIONS: Route-based transport time estimates demonstrate moderate accuracy. These methods can be valuable for informing a host of decisions related to the system organization and patient access to emergency medical care; however, they should be employed with sensitivity to their limitations.
2013
Alagoz, Oguzhan; Chhatwal, Jagpreet; Burnside, Elizabeth S
Optimal Policies for Reducing Unnecessary Follow-up Mammography Exams in Breast Cancer Diagnosis Journal Article
In: Decis Anal, vol. 10, no. 3, pp. 200–224, 2013, ISSN: 1545-8490.
@article{pmid24501588,
title = {Optimal Policies for Reducing Unnecessary Follow-up Mammography Exams in Breast Cancer Diagnosis},
author = {Oguzhan Alagoz and Jagpreet Chhatwal and Elizabeth S Burnside},
doi = {10.1287/deca.2013.0272},
issn = {1545-8490},
year = {2013},
date = {2013-09-01},
journal = {Decis Anal},
volume = {10},
number = {3},
pages = {200--224},
abstract = {Mammography is the most effective screening tool for early diagnosis of breast cancer. Based on the mammography findings, radiologists need to choose from one of the following three alternatives: 1) take immediate diagnostic actions including prompt biopsy to confirm breast cancer; 2) recommend a follow-up mammogram; 3) recommend routine annual mammography. There are no validated structured guidelines based on a decision-analytical framework to aid radiologists in making such patient management decisions. Surprisingly, only 15-45% of the breast biopsies and less than 1% of short-interval follow-up recommendations are found to be malignant, resulting in unnecessary tests and patient-anxiety. We develop a finite-horizon discrete-time Markov decision process (MDP) model that may help radiologists make patient-management decisions to maximize a patient's total expected quality-adjusted life years. We use clinical data to find the policies recommended by the MDP model and also compare them to decisions made by radiologists at a large mammography practice. We also derive the structural properties of the MDP model, including sufficiency conditions that ensure the existence of a double control-limit type policy.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Mammography is the most effective screening tool for early diagnosis of breast cancer. Based on the mammography findings, radiologists need to choose from one of the following three alternatives: 1) take immediate diagnostic actions including prompt biopsy to confirm breast cancer; 2) recommend a follow-up mammogram; 3) recommend routine annual mammography. There are no validated structured guidelines based on a decision-analytical framework to aid radiologists in making such patient management decisions. Surprisingly, only 15-45% of the breast biopsies and less than 1% of short-interval follow-up recommendations are found to be malignant, resulting in unnecessary tests and patient-anxiety. We develop a finite-horizon discrete-time Markov decision process (MDP) model that may help radiologists make patient-management decisions to maximize a patient's total expected quality-adjusted life years. We use clinical data to find the policies recommended by the MDP model and also compare them to decisions made by radiologists at a large mammography practice. We also derive the structural properties of the MDP model, including sufficiency conditions that ensure the existence of a double control-limit type policy.
Ferrante, Shannon Allen; Chhatwal, Jagpreet; Brass, Clifford A; Khoury, Antoine C El; Poordad, Fred; Bronowicki, Jean-Pierre; Elbasha, Elamin H
Boceprevir for previously untreated patients with chronic hepatitis C Genotype 1 infection: a US-based cost-effectiveness modeling study Journal Article
In: BMC Infect Dis, vol. 13, pp. 190, 2013, ISSN: 1471-2334.
@article{pmid23621902,
title = {Boceprevir for previously untreated patients with chronic hepatitis C Genotype 1 infection: a US-based cost-effectiveness modeling study},
author = {Shannon Allen Ferrante and Jagpreet Chhatwal and Clifford A Brass and Antoine C El Khoury and Fred Poordad and Jean-Pierre Bronowicki and Elamin H Elbasha},
doi = {10.1186/1471-2334-13-190},
issn = {1471-2334},
year = {2013},
date = {2013-04-01},
journal = {BMC Infect Dis},
volume = {13},
pages = {190},
abstract = {BACKGROUND: SPRINT-2 demonstrated that boceprevir (BOC), an oral hepatitis C virus (HCV) nonstructural 3 (NS3) protease inhibitor, added to peginterferon alfa-2b (P) and ribavirin (R) significantly increased sustained virologic response rates over PR alone in previously untreated adult patients with chronic HCV genotype 1. We estimated the long-term impact of triple therapy vs. dual therapy on the clinical burden of HCV and performed a cost-effectiveness evaluation.nnMETHODS: A Markov model was used to estimate the incidence of liver complications, discounted costs (2010 US$), quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) of three treatment strategies for treatment-naïve patients with chronic HCV genotype 1. The model simulates the treatment regimens studied in SPRINT-2 in which PR was administered for 4 weeks followed by: 1) placebo plus PR for 44 weeks (PR48); 2) BOC plus PR using response guided therapy (BOC/RGT); and 3) BOC plus PR for 44 weeks (BOC/PR48) and makes projections within and beyond the trial. HCV-related state-transition probabilities, costs, and utilities were obtained from previously published studies. All costs and QALYs were discounted at 3%.nnRESULTS: The model projected approximately 38% and 43% relative reductions in the lifetime incidence of liver complications in the BOC/RGT and BOC/PR48 regimens compared with PR48, respectively. Treatment with BOC/RGT is associated with an incremental cost of $10,348 and an increase of 0.62 QALYs compared to treatment with PR48. Treatment with BOC/PR48 is associated with an incremental cost of $35,727 and an increase of 0.65 QALYs compared to treatment with PR48. The ICERs were $16,792/QALY and $55,162/QALY for the boceprevir-based treatment groups compared with PR48, respectively. The ICER for BOC/PR48 compared with BOC/RGT was $807,804.nnCONCLUSION: The boceprevir-based regimens used in the SPRINT-2 trial were projected to substantially reduce the lifetime incidence of liver complications and increase the QALYs in treatment-naive patients with hepatitis C genotype 1. It was also demonstrated that boceprevir-based regimens offer patients the possibility of experiencing great clinical benefit with a shorter duration of therapy. Both boceprevir-based treatment strategies were projected to be cost-effective at a reasonable threshold in the US when compared to treatment with PR48.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: SPRINT-2 demonstrated that boceprevir (BOC), an oral hepatitis C virus (HCV) nonstructural 3 (NS3) protease inhibitor, added to peginterferon alfa-2b (P) and ribavirin (R) significantly increased sustained virologic response rates over PR alone in previously untreated adult patients with chronic HCV genotype 1. We estimated the long-term impact of triple therapy vs. dual therapy on the clinical burden of HCV and performed a cost-effectiveness evaluation.nnMETHODS: A Markov model was used to estimate the incidence of liver complications, discounted costs (2010 US$), quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) of three treatment strategies for treatment-naïve patients with chronic HCV genotype 1. The model simulates the treatment regimens studied in SPRINT-2 in which PR was administered for 4 weeks followed by: 1) placebo plus PR for 44 weeks (PR48); 2) BOC plus PR using response guided therapy (BOC/RGT); and 3) BOC plus PR for 44 weeks (BOC/PR48) and makes projections within and beyond the trial. HCV-related state-transition probabilities, costs, and utilities were obtained from previously published studies. All costs and QALYs were discounted at 3%.nnRESULTS: The model projected approximately 38% and 43% relative reductions in the lifetime incidence of liver complications in the BOC/RGT and BOC/PR48 regimens compared with PR48, respectively. Treatment with BOC/RGT is associated with an incremental cost of $10,348 and an increase of 0.62 QALYs compared to treatment with PR48. Treatment with BOC/PR48 is associated with an incremental cost of $35,727 and an increase of 0.65 QALYs compared to treatment with PR48. The ICERs were $16,792/QALY and $55,162/QALY for the boceprevir-based treatment groups compared with PR48, respectively. The ICER for BOC/PR48 compared with BOC/RGT was $807,804.nnCONCLUSION: The boceprevir-based regimens used in the SPRINT-2 trial were projected to substantially reduce the lifetime incidence of liver complications and increase the QALYs in treatment-naive patients with hepatitis C genotype 1. It was also demonstrated that boceprevir-based regimens offer patients the possibility of experiencing great clinical benefit with a shorter duration of therapy. Both boceprevir-based treatment strategies were projected to be cost-effective at a reasonable threshold in the US when compared to treatment with PR48.
Elbasha, Elamin H; Chhatwal, Jagpreet; Ferrante, Shannon A; Khoury, Antoine C El; Laires, Pedro A
Cost-effectiveness analysis of boceprevir for the treatment of chronic hepatitis C virus genotype 1 infection in Portugal Journal Article
In: Appl Health Econ Health Policy, vol. 11, no. 1, pp. 65–78, 2013, ISSN: 1179-1896.
@article{pmid23355388,
title = {Cost-effectiveness analysis of boceprevir for the treatment of chronic hepatitis C virus genotype 1 infection in Portugal},
author = {Elamin H Elbasha and Jagpreet Chhatwal and Shannon A Ferrante and Antoine C El Khoury and Pedro A Laires},
doi = {10.1007/s40258-012-0007-8},
issn = {1179-1896},
year = {2013},
date = {2013-02-01},
journal = {Appl Health Econ Health Policy},
volume = {11},
number = {1},
pages = {65--78},
abstract = {BACKGROUND: The recent approval of two protease inhibitors, boceprevir and telaprevir, is likely to change the management of chronic hepatitis C virus (HCV) genotype 1 infection.nnOBJECTIVES: We evaluated the long-term clinical outcomes and the cost effectiveness of therapeutic strategies using boceprevir with peginterferon plus ribavirin (PR) in comparison with PR alone for treating HCV genotype 1 infection in Portugal.nnMETHODS: A Markov model was developed to project the expected lifetime costs and quality-adjusted life-years (QALYs) associated with PR alone and the treatment strategies outlined by the European Medicines Agency in the boceprevir summary of product characteristics. The boceprevir-based therapeutic strategies differ according to whether or not the patient was previously treated and whether or not the patient had compensated cirrhosis. The model simulated the experience of a series of cohorts of chronically HCV-infected patients (each defined by age, sex, race and fibrosis score). All treatment-related inputs were obtained from boceprevir clinical trials - SPRINT-2, RESPOND-2 and PROVIDE. Estimates of the natural history parameters and health state utilities were based on published studies. Portugal-specific annual direct costs of HCV health states were estimated by convening a panel of experts to derive health state resource use and multiplying the results by national unit costs. The model was developed from a healthcare system perspective with a timeframe corresponding to the remaining duration of the patients' lifetimes. Both future costs and QALYs were discounted at 5 %. To test the robustness of the conclusions, we conducted deterministic and probabilistic sensitivity analyses.nnRESULTS: In comparison with the treatment with PR alone, boceprevir-based regimens were projected to reduce the lifetime incidence of advanced liver disease, liver transplantation, and liver-related death by 45-51 % and increase life expectancy by 2.3-4.3 years. Although the addition of BOC increased treatment costs by €13,300-€19,700, the reduction of disease burden resulted in a decrease of €5,400-€9,000 in discounted health state costs and an increase of 0.68-1.23 in discounted QALYs per patient. The incremental cost-effectiveness ratios of the boceprevir-based regimens compared with PR among previously untreated and previously treated patients were €11,600/QALY and €8,700/QALY, respectively. The results were most sensitive to variations in sustained virologic response rates, discount rates and age at treatment.nnCONCLUSIONS: Adding boceprevir to PR was projected to reduce the number of liver complications and liver-related deaths, and to be cost effective in treating both previously untreated and treated patients.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: The recent approval of two protease inhibitors, boceprevir and telaprevir, is likely to change the management of chronic hepatitis C virus (HCV) genotype 1 infection.nnOBJECTIVES: We evaluated the long-term clinical outcomes and the cost effectiveness of therapeutic strategies using boceprevir with peginterferon plus ribavirin (PR) in comparison with PR alone for treating HCV genotype 1 infection in Portugal.nnMETHODS: A Markov model was developed to project the expected lifetime costs and quality-adjusted life-years (QALYs) associated with PR alone and the treatment strategies outlined by the European Medicines Agency in the boceprevir summary of product characteristics. The boceprevir-based therapeutic strategies differ according to whether or not the patient was previously treated and whether or not the patient had compensated cirrhosis. The model simulated the experience of a series of cohorts of chronically HCV-infected patients (each defined by age, sex, race and fibrosis score). All treatment-related inputs were obtained from boceprevir clinical trials - SPRINT-2, RESPOND-2 and PROVIDE. Estimates of the natural history parameters and health state utilities were based on published studies. Portugal-specific annual direct costs of HCV health states were estimated by convening a panel of experts to derive health state resource use and multiplying the results by national unit costs. The model was developed from a healthcare system perspective with a timeframe corresponding to the remaining duration of the patients' lifetimes. Both future costs and QALYs were discounted at 5 %. To test the robustness of the conclusions, we conducted deterministic and probabilistic sensitivity analyses.nnRESULTS: In comparison with the treatment with PR alone, boceprevir-based regimens were projected to reduce the lifetime incidence of advanced liver disease, liver transplantation, and liver-related death by 45-51 % and increase life expectancy by 2.3-4.3 years. Although the addition of BOC increased treatment costs by €13,300-€19,700, the reduction of disease burden resulted in a decrease of €5,400-€9,000 in discounted health state costs and an increase of 0.68-1.23 in discounted QALYs per patient. The incremental cost-effectiveness ratios of the boceprevir-based regimens compared with PR among previously untreated and previously treated patients were €11,600/QALY and €8,700/QALY, respectively. The results were most sensitive to variations in sustained virologic response rates, discount rates and age at treatment.nnCONCLUSIONS: Adding boceprevir to PR was projected to reduce the number of liver complications and liver-related deaths, and to be cost effective in treating both previously untreated and treated patients.
Odhiambo, Raymond; Chhatwal, Jagpreet; Ferrante, Shannon Allen; Khoury, Antoine El; Elbasha, Elamin
Economic Evaluation of Boceprevir for the Treatment of Patients with Genotype 1 Chronic Hepatitis C Virus Infection in Hungary Journal Article
In: J Health Econ Outcomes Res, vol. 1, no. 1, pp. 62–82, 2013, ISSN: 2327-2236.
@article{pmid37664146,
title = {Economic Evaluation of Boceprevir for the Treatment of Patients with Genotype 1 Chronic Hepatitis C Virus Infection in Hungary},
author = {Raymond Odhiambo and Jagpreet Chhatwal and Shannon Allen Ferrante and Antoine El Khoury and Elamin Elbasha},
doi = {10.36469/9854},
issn = {2327-2236},
year = {2013},
date = {2013-01-01},
journal = {J Health Econ Outcomes Res},
volume = {1},
number = {1},
pages = {62--82},
abstract = { Recent international, randomized, placebo-controlled clinical trials (SPRINT-2; RESPOND-2) demonstrated that the triple combination of peginterferon (PEG), ribavirin (RBV) and boceprevir (BOC) was more efficacious than the standard dual therapy of PEG and RBV in treatment of patients chronically infected with genotype 1 hepatitis C virus (HCV) infection. The objective of this study was to evaluate the cost-effectiveness of triple therapy in both treatment-naive and treatment-experienced patients in Hungary. A Markov model was developed to evaluate the long-term clinical benefits and the costeffectiveness of the triple therapy from the Hungarian payer perspective. Model states were fibrosis (F0-F4, defined using METAVIR fibrosis scores), decompensated cirrhosis (DC), hepatocellular carcinoma (HCC), liver transplantation (LT), and liver-related deaths (LD). Efficacy was estimated from SPRINT-2 and RESPOND-2 studies. Disease progression rates and health state utilities used in the model were obtained from published studies. Estimates of probability of liver transplantation and cost were based on an analysis of the Hungarian Sick Fund database. All cost and benefits were discounted at 5% per year. Compared to dual therapy, triple therapy was projected to increase the life expectancy by 0.98 and 2.42 life years and increase the quality-adjusted life years (QALY) by 0.59 and 1.13 in treatment-naive and treatment-experienced patients, respectively. The corresponding incremental cost-effectiveness ratios were HUF7,747,962 (€26,717) and HUF5,888,240 (€20,304) per QALY. The lifetime incidence of severe liver disease events (DC, HCC, LT, LD) were projected to decrease by 45% and 61% in treatment-naïve and treatment-experienced patients treated with triple therapy groups in comparison with PEG-RBV treatment. The addition of boceprevir to standard therapy for the treatment of patients with genotype 1 chronic HCV infection in Hungary is projected to be cost-effective using a commonly used willingness to pay threshold of HUF 8.46 million (3 times gross domestic product per capita).},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Recent international, randomized, placebo-controlled clinical trials (SPRINT-2; RESPOND-2) demonstrated that the triple combination of peginterferon (PEG), ribavirin (RBV) and boceprevir (BOC) was more efficacious than the standard dual therapy of PEG and RBV in treatment of patients chronically infected with genotype 1 hepatitis C virus (HCV) infection. The objective of this study was to evaluate the cost-effectiveness of triple therapy in both treatment-naive and treatment-experienced patients in Hungary. A Markov model was developed to evaluate the long-term clinical benefits and the costeffectiveness of the triple therapy from the Hungarian payer perspective. Model states were fibrosis (F0-F4, defined using METAVIR fibrosis scores), decompensated cirrhosis (DC), hepatocellular carcinoma (HCC), liver transplantation (LT), and liver-related deaths (LD). Efficacy was estimated from SPRINT-2 and RESPOND-2 studies. Disease progression rates and health state utilities used in the model were obtained from published studies. Estimates of probability of liver transplantation and cost were based on an analysis of the Hungarian Sick Fund database. All cost and benefits were discounted at 5% per year. Compared to dual therapy, triple therapy was projected to increase the life expectancy by 0.98 and 2.42 life years and increase the quality-adjusted life years (QALY) by 0.59 and 1.13 in treatment-naive and treatment-experienced patients, respectively. The corresponding incremental cost-effectiveness ratios were HUF7,747,962 (€26,717) and HUF5,888,240 (€20,304) per QALY. The lifetime incidence of severe liver disease events (DC, HCC, LT, LD) were projected to decrease by 45% and 61% in treatment-naïve and treatment-experienced patients treated with triple therapy groups in comparison with PEG-RBV treatment. The addition of boceprevir to standard therapy for the treatment of patients with genotype 1 chronic HCV infection in Hungary is projected to be cost-effective using a commonly used willingness to pay threshold of HUF 8.46 million (3 times gross domestic product per capita).
Chhatwal, Jagpreet; Ferrante, Shannon A; Brass, Cliff; Khoury, Antoine C El; Burroughs, Margaret; Bacon, Bruce; Esteban-Mur, Rafael; Elbasha, Elamin H
Cost-effectiveness of boceprevir in patients previously treated for chronic hepatitis C genotype 1 infection in the United States Journal Article
In: Value Health, vol. 16, no. 6, pp. 973–986, 2013, ISSN: 1524-4733.
@article{pmid24041347,
title = {Cost-effectiveness of boceprevir in patients previously treated for chronic hepatitis C genotype 1 infection in the United States},
author = {Jagpreet Chhatwal and Shannon A Ferrante and Cliff Brass and Antoine C El Khoury and Margaret Burroughs and Bruce Bacon and Rafael Esteban-Mur and Elamin H Elbasha},
doi = {10.1016/j.jval.2013.07.006},
issn = {1524-4733},
year = {2013},
date = {2013-01-01},
journal = {Value Health},
volume = {16},
number = {6},
pages = {973--986},
abstract = {OBJECTIVES: The phase 3 trial, Serine Protease Inhibitor Boceprevir and PegIntron/Rebetol-2 (RESPOND-2), demonstrated that the addition of boceprevir (BOC) to peginterferon-ribavirin (PR) resulted in significantly higher rates of sustained virologic response (SVR) in previously treated patients with chronic hepatitis C virus (HCV) genotype-1 infection as compared with PR alone. We evaluated the cost-effectiveness of treatment with BOC in previously treated patients with chronic hepatitis C in the United States using treatment-related data from RESPOND-2 and PROVIDE studies.nnMETHODS: We developed a Markov cohort model to project the burden of HCV disease, lifetime costs, and quality-adjusted life-years associated with PR and two BOC-based therapies-response-guided therapy (BOC/RGT) and fixed-duration therapy for 48 weeks (BOC/PR48). We estimated treatment-related inputs (efficacy, adverse events, and discontinuations) from clinical trials and obtained disease progression rates, costs, and quality-of-life data from published studies. We estimated the incremental cost-effectiveness ratio (ICER) for BOC-based regimens as studied in RESPOND-2, as well as by patient's prior response to treatment and the IL-28B genotype.nnRESULTS: BOC-based regimens were projected to reduce the lifetime incidence of liver-related complications by 43% to 53% in comparison with treatment with PR. The ICER of BOC/RGT in comparison with that of PR was $30,200, and the ICER of BOC/PR48 in comparison with that of BOC/RGT was $91,500. At a willingness-to-pay threshold of $50,000, the probabilities of BOC/RGT and BOC/PR48 being the preferred option were 0.74 and 0.25, respectively.nnCONCLUSIONS: In patients previously treated for chronic HCV genotype-1 infection, BOC was projected to increase quality-adjusted life-years and reduce the lifetime incidence of liver complications. In addition, BOC-based therapies were projected to be cost-effective in comparison with PR alone at commonly used willingness-to-pay thresholds.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
OBJECTIVES: The phase 3 trial, Serine Protease Inhibitor Boceprevir and PegIntron/Rebetol-2 (RESPOND-2), demonstrated that the addition of boceprevir (BOC) to peginterferon-ribavirin (PR) resulted in significantly higher rates of sustained virologic response (SVR) in previously treated patients with chronic hepatitis C virus (HCV) genotype-1 infection as compared with PR alone. We evaluated the cost-effectiveness of treatment with BOC in previously treated patients with chronic hepatitis C in the United States using treatment-related data from RESPOND-2 and PROVIDE studies.nnMETHODS: We developed a Markov cohort model to project the burden of HCV disease, lifetime costs, and quality-adjusted life-years associated with PR and two BOC-based therapies-response-guided therapy (BOC/RGT) and fixed-duration therapy for 48 weeks (BOC/PR48). We estimated treatment-related inputs (efficacy, adverse events, and discontinuations) from clinical trials and obtained disease progression rates, costs, and quality-of-life data from published studies. We estimated the incremental cost-effectiveness ratio (ICER) for BOC-based regimens as studied in RESPOND-2, as well as by patient's prior response to treatment and the IL-28B genotype.nnRESULTS: BOC-based regimens were projected to reduce the lifetime incidence of liver-related complications by 43% to 53% in comparison with treatment with PR. The ICER of BOC/RGT in comparison with that of PR was $30,200, and the ICER of BOC/PR48 in comparison with that of BOC/RGT was $91,500. At a willingness-to-pay threshold of $50,000, the probabilities of BOC/RGT and BOC/PR48 being the preferred option were 0.74 and 0.25, respectively.nnCONCLUSIONS: In patients previously treated for chronic HCV genotype-1 infection, BOC was projected to increase quality-adjusted life-years and reduce the lifetime incidence of liver complications. In addition, BOC-based therapies were projected to be cost-effective in comparison with PR alone at commonly used willingness-to-pay thresholds.
Luangkesorn, Kiatikun Louis; Ghiasabadi, Farhad; Chhatwal, Jagpreet
A sequential experimental design method to evaluate a combination of school closure and vaccination policies to control an H1N1-like pandemic Journal Article
In: J Public Health Manag Pract, vol. 19 Suppl 2, no. 0 2, pp. S37–S41, 2013, ISSN: 1550-5022.
@article{pmid23903393,
title = {A sequential experimental design method to evaluate a combination of school closure and vaccination policies to control an H1N1-like pandemic},
author = {Kiatikun Louis Luangkesorn and Farhad Ghiasabadi and Jagpreet Chhatwal},
doi = {10.1097/PHH.0b013e3182939a5c},
issn = {1550-5022},
year = {2013},
date = {2013-01-01},
journal = {J Public Health Manag Pract},
volume = {19 Suppl 2},
number = {0 2},
pages = {S37--S41},
abstract = {CONTEXT: During the 2009 H1N1 pandemic, computational agent-based models (ABMs) were extensively used to evaluate interventions to control the spread of emerging pathogens. However, evaluating different possible combinations of interventions using ABMs can be computationally very expensive and time-consuming. Therefore, most policy studies have examined the impact of a single policy decision.nnOBJECTIVE: To apply a sequential experimental design method with an ABM to analyze policy alternatives composed of a combination of school closure and vaccination policies to provide a set of promising "optimal" combinations of policies to control an H1N1-type epidemic to policy makers.nnMETHODS: We used an open-source agent-based modeling system, FRED (A Framework for Reconstructing Epidemiological Dynamic), to simulate the spread of an H1N1 epidemic in Alleghany County, Pennsylvania, with a census-based synthetic population. We used an approach called best subset selection method to evaluate 72 alternative policies consisting of a combination of options for school closure threshold, closure duration, Advisory Committee on Immunization Practices prioritization, and second-dose vaccination prioritization policies. Using the attack rate as a performance measure, best subset selection enabled us to eliminate inferior alternatives and identify a small group of alternative policies that could be further evaluated on the basis of other criteria.nnRESULTS: Our sequential design approach to evaluate a combination of alternative mitigation policies leads to a savings in computational effort by a factor of 2 when examining combinations of school closure and vaccination policies.nnCONCLUSIONS: Best subset selection demonstrates a substantial reduction in the computational burden of a large-scale ABM in evaluating several alternative policies. Our method also provides policy makers with a set of promising policy combinations for further evaluation based on implementation considerations or other criteria.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
CONTEXT: During the 2009 H1N1 pandemic, computational agent-based models (ABMs) were extensively used to evaluate interventions to control the spread of emerging pathogens. However, evaluating different possible combinations of interventions using ABMs can be computationally very expensive and time-consuming. Therefore, most policy studies have examined the impact of a single policy decision.nnOBJECTIVE: To apply a sequential experimental design method with an ABM to analyze policy alternatives composed of a combination of school closure and vaccination policies to provide a set of promising "optimal" combinations of policies to control an H1N1-type epidemic to policy makers.nnMETHODS: We used an open-source agent-based modeling system, FRED (A Framework for Reconstructing Epidemiological Dynamic), to simulate the spread of an H1N1 epidemic in Alleghany County, Pennsylvania, with a census-based synthetic population. We used an approach called best subset selection method to evaluate 72 alternative policies consisting of a combination of options for school closure threshold, closure duration, Advisory Committee on Immunization Practices prioritization, and second-dose vaccination prioritization policies. Using the attack rate as a performance measure, best subset selection enabled us to eliminate inferior alternatives and identify a small group of alternative policies that could be further evaluated on the basis of other criteria.nnRESULTS: Our sequential design approach to evaluate a combination of alternative mitigation policies leads to a savings in computational effort by a factor of 2 when examining combinations of school closure and vaccination policies.nnCONCLUSIONS: Best subset selection demonstrates a substantial reduction in the computational burden of a large-scale ABM in evaluating several alternative policies. Our method also provides policy makers with a set of promising policy combinations for further evaluation based on implementation considerations or other criteria.
2012
Burnside, Elizabeth S; Chhatwal, Jagpreet; Alagoz, Oguzhan
What is the optimal threshold at which to recommend breast biopsy? Journal Article
In: PLoS One, vol. 7, no. 11, pp. e48820, 2012, ISSN: 1932-6203.
@article{pmid23144986,
title = {What is the optimal threshold at which to recommend breast biopsy?},
author = {Elizabeth S Burnside and Jagpreet Chhatwal and Oguzhan Alagoz},
doi = {10.1371/journal.pone.0048820},
issn = {1932-6203},
year = {2012},
date = {2012-01-01},
journal = {PLoS One},
volume = {7},
number = {11},
pages = {e48820},
abstract = {BACKGROUND: A 2% threshold, traditionally used as a level above which breast biopsy recommended, has been generalized to all patients from several specific situations analyzed in the literature. We use a sequential decision analytic model considering clinical and mammography features to determine the optimal general threshold for image guided breast biopsy and the sensitivity of this threshold to variation of these features.nnMETHODOLOGY/PRINCIPAL FINDINGS: We built a decision analytical model called a Markov Decision Process (MDP) model, which determines the optimal threshold of breast cancer risk to perform breast biopsy in order to maximize a patient's total quality-adjusted life years (QALYs). The optimal biopsy threshold is determined based on a patient's probability of breast cancer estimated by a logistic regression model (LRM) which uses demographic risk factors (age, family history, and hormone use) and mammographic findings (described using the established lexicon-BI-RADS). We estimate the MDP model's parameters using SEER data (prevalence of invasive vs. in situ disease, stage at diagnosis, and survival), US life tables (all cause mortality), and the medical literature (biopsy disutility and treatment efficacy) to determine the optimal "base case" risk threshold for breast biopsy and perform sensitivity analysis. The base case MDP model reveals that 2% is the optimal threshold for breast biopsy for patients between 42 and 75 however the thresholds below age 42 is lower (1%) and above age 75 is higher (range of 3-5%). Our sensitivity analysis reveals that the optimal biopsy threshold varies most notably with changes in age and disutility of biopsy.nnCONCLUSIONS/SIGNIFICANCE: Our MDP model validates the 2% threshold currently used for biopsy but shows this optimal threshold varies substantially with patient age and biopsy disutility.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: A 2% threshold, traditionally used as a level above which breast biopsy recommended, has been generalized to all patients from several specific situations analyzed in the literature. We use a sequential decision analytic model considering clinical and mammography features to determine the optimal general threshold for image guided breast biopsy and the sensitivity of this threshold to variation of these features.nnMETHODOLOGY/PRINCIPAL FINDINGS: We built a decision analytical model called a Markov Decision Process (MDP) model, which determines the optimal threshold of breast cancer risk to perform breast biopsy in order to maximize a patient's total quality-adjusted life years (QALYs). The optimal biopsy threshold is determined based on a patient's probability of breast cancer estimated by a logistic regression model (LRM) which uses demographic risk factors (age, family history, and hormone use) and mammographic findings (described using the established lexicon-BI-RADS). We estimate the MDP model's parameters using SEER data (prevalence of invasive vs. in situ disease, stage at diagnosis, and survival), US life tables (all cause mortality), and the medical literature (biopsy disutility and treatment efficacy) to determine the optimal "base case" risk threshold for breast biopsy and perform sensitivity analysis. The base case MDP model reveals that 2% is the optimal threshold for breast biopsy for patients between 42 and 75 however the thresholds below age 42 is lower (1%) and above age 75 is higher (range of 3-5%). Our sensitivity analysis reveals that the optimal biopsy threshold varies most notably with changes in age and disutility of biopsy.nnCONCLUSIONS/SIGNIFICANCE: Our MDP model validates the 2% threshold currently used for biopsy but shows this optimal threshold varies substantially with patient age and biopsy disutility.
2010
Chhatwal, Jagpreet; Alagoz, Oguzhan; Burnside, Elizabeth S
Optimal Breast Biopsy Decision-Making Based on Mammographic Features and Demographic Factors Journal Article
In: Oper Res, vol. 58, no. 6, pp. 1577–1591, 2010, ISSN: 0030-364X.
@article{pmid21415931,
title = {Optimal Breast Biopsy Decision-Making Based on Mammographic Features and Demographic Factors},
author = {Jagpreet Chhatwal and Oguzhan Alagoz and Elizabeth S Burnside},
doi = {10.1287/opre.1100.0877},
issn = {0030-364X},
year = {2010},
date = {2010-11-01},
journal = {Oper Res},
volume = {58},
number = {6},
pages = {1577--1591},
abstract = {Breast cancer is the most common non-skin cancer affecting women in the United States, where every year more than 20 million mammograms are performed. Breast biopsy is commonly performed on the suspicious findings on mammograms to confirm the presence of cancer. Currently, 700,000 biopsies are performed annually in the U.S.; 55%-85% of these biopsies ultimately are found to be benign breast lesions, resulting in unnecessary treatments, patient anxiety, and expenditures. This paper addresses the decision problem faced by radiologists: When should a woman be sent for biopsy based on her mammographic features and demographic factors? This problem is formulated as a finite-horizon discrete-time Markov decision process. The optimal policy of our model shows that the decision to biopsy should take the age of patient into account; particularly, an older patient's risk threshold for biopsy should be higher than that of a younger patient. When applied to the clinical data, our model outperforms radiologists in the biopsy decision-making problem. This study also derives structural properties of the model, including sufficiency conditions that ensure the existence of a control-limit type policy and nondecreasing control-limits with age.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Breast cancer is the most common non-skin cancer affecting women in the United States, where every year more than 20 million mammograms are performed. Breast biopsy is commonly performed on the suspicious findings on mammograms to confirm the presence of cancer. Currently, 700,000 biopsies are performed annually in the U.S.; 55%-85% of these biopsies ultimately are found to be benign breast lesions, resulting in unnecessary treatments, patient anxiety, and expenditures. This paper addresses the decision problem faced by radiologists: When should a woman be sent for biopsy based on her mammographic features and demographic factors? This problem is formulated as a finite-horizon discrete-time Markov decision process. The optimal policy of our model shows that the decision to biopsy should take the age of patient into account; particularly, an older patient's risk threshold for biopsy should be higher than that of a younger patient. When applied to the clinical data, our model outperforms radiologists in the biopsy decision-making problem. This study also derives structural properties of the model, including sufficiency conditions that ensure the existence of a control-limit type policy and nondecreasing control-limits with age.
Ayer, Turgay; Alagoz, Oguzhan; Chhatwal, Jagpreet; Shavlik, Jude W; Kahn, Charles E; Burnside, Elizabeth S
Breast cancer risk estimation with artificial neural networks revisited: discrimination and calibration Journal Article
In: Cancer, vol. 116, no. 14, pp. 3310–3321, 2010, ISSN: 0008-543X.
@article{pmid20564067,
title = {Breast cancer risk estimation with artificial neural networks revisited: discrimination and calibration},
author = {Turgay Ayer and Oguzhan Alagoz and Jagpreet Chhatwal and Jude W Shavlik and Charles E Kahn and Elizabeth S Burnside},
doi = {10.1002/cncr.25081},
issn = {0008-543X},
year = {2010},
date = {2010-07-01},
journal = {Cancer},
volume = {116},
number = {14},
pages = {3310--3321},
abstract = {BACKGROUND: Discriminating malignant breast lesions from benign ones and accurately predicting the risk of breast cancer for individual patients are crucial to successful clinical decisions. In the past, several artificial neural network (ANN) models have been developed for breast cancer-risk prediction. All studies have reported discrimination performance, but not one has assessed calibration, which is an equivalently important measure for accurate risk prediction. In this study, the authors have evaluated whether an artificial neural network (ANN) trained on a large prospectively collected dataset of consecutive mammography findings can discriminate between benign and malignant disease and accurately predict the probability of breast cancer for individual patients.nnMETHODS: Our dataset consisted of 62,219 consecutively collected mammography findings matched with the Wisconsin State Cancer Reporting System. The authors built a 3-layer feedforward ANN with 1000 hidden-layer nodes. The authors trained and tested their ANN by using 10-fold cross-validation to predict the risk of breast cancer. The authors used area the under the receiver-operating characteristic curve (AUC), sensitivity, and specificity to evaluate discriminative performance of the radiologists and their ANN. The authors assessed the accuracy of risk prediction (ie, calibration) of their ANN by using the Hosmer-Lemeshow (H-L) goodness-of-fit test.nnRESULTS: Their ANN demonstrated superior discrimination (AUC, 0.965) compared with the radiologists (AUC, 0.939; P<.001). The authors' ANN was also well calibrated as shown by an H-L goodness of fit P-value of .13.nnCONCLUSIONS: The authors' ANN can effectively discriminate malignant abnormalities from benign ones and accurately predict the risk of breast cancer for individual abnormalities.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND: Discriminating malignant breast lesions from benign ones and accurately predicting the risk of breast cancer for individual patients are crucial to successful clinical decisions. In the past, several artificial neural network (ANN) models have been developed for breast cancer-risk prediction. All studies have reported discrimination performance, but not one has assessed calibration, which is an equivalently important measure for accurate risk prediction. In this study, the authors have evaluated whether an artificial neural network (ANN) trained on a large prospectively collected dataset of consecutive mammography findings can discriminate between benign and malignant disease and accurately predict the probability of breast cancer for individual patients.nnMETHODS: Our dataset consisted of 62,219 consecutively collected mammography findings matched with the Wisconsin State Cancer Reporting System. The authors built a 3-layer feedforward ANN with 1000 hidden-layer nodes. The authors trained and tested their ANN by using 10-fold cross-validation to predict the risk of breast cancer. The authors used area the under the receiver-operating characteristic curve (AUC), sensitivity, and specificity to evaluate discriminative performance of the radiologists and their ANN. The authors assessed the accuracy of risk prediction (ie, calibration) of their ANN by using the Hosmer-Lemeshow (H-L) goodness-of-fit test.nnRESULTS: Their ANN demonstrated superior discrimination (AUC, 0.965) compared with the radiologists (AUC, 0.939; P<.001). The authors' ANN was also well calibrated as shown by an H-L goodness of fit P-value of .13.nnCONCLUSIONS: The authors' ANN can effectively discriminate malignant abnormalities from benign ones and accurately predict the risk of breast cancer for individual abnormalities.
Ayer, Turgay; Chhatwal, Jagpreet; Alagoz, Oguzhan; Kahn, Charles E; Woods, Ryan W; Burnside, Elizabeth S
Informatics in radiology: comparison of logistic regression and artificial neural network models in breast cancer risk estimation Journal Article
In: Radiographics, vol. 30, no. 1, pp. 13–22, 2010, ISSN: 1527-1323.
@article{pmid19901087,
title = {Informatics in radiology: comparison of logistic regression and artificial neural network models in breast cancer risk estimation},
author = {Turgay Ayer and Jagpreet Chhatwal and Oguzhan Alagoz and Charles E Kahn and Ryan W Woods and Elizabeth S Burnside},
doi = {10.1148/rg.301095057},
issn = {1527-1323},
year = {2010},
date = {2010-01-01},
journal = {Radiographics},
volume = {30},
number = {1},
pages = {13--22},
abstract = {Computer models in medical diagnosis are being developed to help physicians differentiate between healthy patients and patients with disease. These models can aid in successful decision making by allowing calculation of disease likelihood on the basis of known patient characteristics and clinical test results. Two of the most frequently used computer models in clinical risk estimation are logistic regression and an artificial neural network. A study was conducted to review and compare these two models, elucidate the advantages and disadvantages of each, and provide criteria for model selection. The two models were used for estimation of breast cancer risk on the basis of mammographic descriptors and demographic risk factors. Although they demonstrated similar performance, the two models have unique characteristics-strengths as well as limitations-that must be considered and may prove complementary in contributing to improved clinical decision making.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Computer models in medical diagnosis are being developed to help physicians differentiate between healthy patients and patients with disease. These models can aid in successful decision making by allowing calculation of disease likelihood on the basis of known patient characteristics and clinical test results. Two of the most frequently used computer models in clinical risk estimation are logistic regression and an artificial neural network. A study was conducted to review and compare these two models, elucidate the advantages and disadvantages of each, and provide criteria for model selection. The two models were used for estimation of breast cancer risk on the basis of mammographic descriptors and demographic risk factors. Although they demonstrated similar performance, the two models have unique characteristics-strengths as well as limitations-that must be considered and may prove complementary in contributing to improved clinical decision making.
2009
Burnside, Elizabeth S; Davis, Jesse; Chhatwal, Jagpreet; Alagoz, Oguzhan; Lindstrom, Mary J; Geller, Berta M; Littenberg, Benjamin; Shaffer, Katherine A; Kahn, Charles E; Page, C David
Probabilistic computer model developed from clinical data in national mammography database format to classify mammographic findings Journal Article
In: Radiology, vol. 251, no. 3, pp. 663–672, 2009, ISSN: 1527-1315.
@article{pmid19366902,
title = {Probabilistic computer model developed from clinical data in national mammography database format to classify mammographic findings},
author = {Elizabeth S Burnside and Jesse Davis and Jagpreet Chhatwal and Oguzhan Alagoz and Mary J Lindstrom and Berta M Geller and Benjamin Littenberg and Katherine A Shaffer and Charles E Kahn and C David Page},
doi = {10.1148/radiol.2513081346},
issn = {1527-1315},
year = {2009},
date = {2009-06-01},
journal = {Radiology},
volume = {251},
number = {3},
pages = {663--672},
abstract = {PURPOSE: To determine whether a Bayesian network trained on a large database of patient demographic risk factors and radiologist-observed findings from consecutive clinical mammography examinations can exceed radiologist performance in the classification of mammographic findings as benign or malignant.nnMATERIALS AND METHODS: The institutional review board exempted this HIPAA-compliant retrospective study from requiring informed consent. Structured reports from 48 744 consecutive pooled screening and diagnostic mammography examinations in 18 269 patients from April 5, 1999 to February 9, 2004 were collected. Mammographic findings were matched with a state cancer registry, which served as the reference standard. By using 10-fold cross validation, the Bayesian network was tested and trained to estimate breast cancer risk by using demographic risk factors (age, family and personal history of breast cancer, and use of hormone replacement therapy) and mammographic findings recorded in the Breast Imaging Reporting and Data System lexicon. The performance of radiologists compared with the Bayesian network was evaluated by using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity.nnRESULTS: The Bayesian network significantly exceeded the performance of interpreting radiologists in terms of AUC (0.960 vs 0.939, P = .002), sensitivity (90.0% vs 85.3%, P < .001), and specificity (93.0% vs 88.1%, P < .001).nnCONCLUSION: On the basis of prospectively collected variables, the evaluated Bayesian network can predict the probability of breast cancer and exceed interpreting radiologist performance. Bayesian networks may help radiologists improve mammographic interpretation.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
PURPOSE: To determine whether a Bayesian network trained on a large database of patient demographic risk factors and radiologist-observed findings from consecutive clinical mammography examinations can exceed radiologist performance in the classification of mammographic findings as benign or malignant.nnMATERIALS AND METHODS: The institutional review board exempted this HIPAA-compliant retrospective study from requiring informed consent. Structured reports from 48 744 consecutive pooled screening and diagnostic mammography examinations in 18 269 patients from April 5, 1999 to February 9, 2004 were collected. Mammographic findings were matched with a state cancer registry, which served as the reference standard. By using 10-fold cross validation, the Bayesian network was tested and trained to estimate breast cancer risk by using demographic risk factors (age, family and personal history of breast cancer, and use of hormone replacement therapy) and mammographic findings recorded in the Breast Imaging Reporting and Data System lexicon. The performance of radiologists compared with the Bayesian network was evaluated by using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity.nnRESULTS: The Bayesian network significantly exceeded the performance of interpreting radiologists in terms of AUC (0.960 vs 0.939, P = .002), sensitivity (90.0% vs 85.3%, P < .001), and specificity (93.0% vs 88.1%, P < .001).nnCONCLUSION: On the basis of prospectively collected variables, the evaluated Bayesian network can predict the probability of breast cancer and exceed interpreting radiologist performance. Bayesian networks may help radiologists improve mammographic interpretation.
Chhatwal, Jagpreet; Alagoz, Oguzhan; Lindstrom, Mary J; Kahn, Charles E; Shaffer, Katherine A; Burnside, Elizabeth S
A logistic regression model based on the national mammography database format to aid breast cancer diagnosis Journal Article
In: AJR Am J Roentgenol, vol. 192, no. 4, pp. 1117–1127, 2009, ISSN: 1546-3141.
@article{pmid19304723,
title = {A logistic regression model based on the national mammography database format to aid breast cancer diagnosis},
author = {Jagpreet Chhatwal and Oguzhan Alagoz and Mary J Lindstrom and Charles E Kahn and Katherine A Shaffer and Elizabeth S Burnside},
doi = {10.2214/AJR.07.3345},
issn = {1546-3141},
year = {2009},
date = {2009-04-01},
journal = {AJR Am J Roentgenol},
volume = {192},
number = {4},
pages = {1117--1127},
abstract = {OBJECTIVE: The purpose of our study was to create a breast cancer risk estimation model based on the descriptors of the National Mammography Database using logistic regression that can aid in decision making for the early detection of breast cancer.nnMATERIALS AND METHODS: We created two logistic regression models based on the mammography features and demographic data for 62,219 consecutive mammography records from 48,744 studies in 18,269 [corrected] patients reported using the Breast Imaging Reporting and Data System (BI-RADS) lexicon and the National Mammography Database format between April 5, 1999 and February 9, 2004. State cancer registry outcomes matched with our data served as the reference standard. The probability of cancer was the outcome in both models. Model 2 was built using all variables in Model 1 plus radiologists' BI-RADS assessment categories. We used 10-fold cross-validation to train and test the model and to calculate the area under the receiver operating characteristic curves (A(z)) to measure the performance. Both models were compared with the radiologists' BI-RADS assessments.nnRESULTS: Radiologists achieved an A(z) value of 0.939 +/- 0.011. The A(z) was 0.927 +/- 0.015 for Model 1 and 0.963 +/- 0.009 for Model 2. At 90% specificity, the sensitivity of Model 2 (90%) was significantly better (p < 0.001) than that of radiologists (82%) and Model 1 (83%). At 85% sensitivity, the specificity of Model 2 (96%) was significantly better (p < 0.001) than that of radiologists (88%) and Model 1 (87%).nnCONCLUSION: Our logistic regression model can effectively discriminate between benign and malignant breast disease and can identify the most important features associated with breast cancer.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
OBJECTIVE: The purpose of our study was to create a breast cancer risk estimation model based on the descriptors of the National Mammography Database using logistic regression that can aid in decision making for the early detection of breast cancer.nnMATERIALS AND METHODS: We created two logistic regression models based on the mammography features and demographic data for 62,219 consecutive mammography records from 48,744 studies in 18,269 [corrected] patients reported using the Breast Imaging Reporting and Data System (BI-RADS) lexicon and the National Mammography Database format between April 5, 1999 and February 9, 2004. State cancer registry outcomes matched with our data served as the reference standard. The probability of cancer was the outcome in both models. Model 2 was built using all variables in Model 1 plus radiologists' BI-RADS assessment categories. We used 10-fold cross-validation to train and test the model and to calculate the area under the receiver operating characteristic curves (A(z)) to measure the performance. Both models were compared with the radiologists' BI-RADS assessments.nnRESULTS: Radiologists achieved an A(z) value of 0.939 +/- 0.011. The A(z) was 0.927 +/- 0.015 for Model 1 and 0.963 +/- 0.009 for Model 2. At 90% specificity, the sensitivity of Model 2 (90%) was significantly better (p < 0.001) than that of radiologists (82%) and Model 1 (83%). At 85% sensitivity, the specificity of Model 2 (96%) was significantly better (p < 0.001) than that of radiologists (88%) and Model 1 (87%).nnCONCLUSION: Our logistic regression model can effectively discriminate between benign and malignant breast disease and can identify the most important features associated with breast cancer.