Physician Workforce, Primary Care and 40 V.K.Chetty, Life - - PowerPoint PPT Presentation

▶
physician workforce primary care and
SMART_READER_LITE
LIVE PREVIEW

Physician Workforce, Primary Care and 40 V.K.Chetty, Life - - PowerPoint PPT Presentation

Dec 02, 2022 232 likes •480 views

Physician Workforce, Primary Care and Life Expectancy at Physician Workforce, Primary Care and 40 V.K.Chetty, Life Expectancy at 40 Benjamin Petrie and Larry Culpepper V.K.Chetty, Benjamin Petrie and Larry Culpepper February,2018

slide-1
SLIDE 1

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Physician Workforce, Primary Care and Life Expectancy at 40

V.K.Chetty, Benjamin Petrie and Larry Culpepper February,2018

slide-2
SLIDE 2

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Raj Chetty,...David Cutler JAMA study

The Association Between Income and Life Expectancy in the United States, 2001-2014 JAMA 2016

slide-3
SLIDE 3

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Raj Chetty,...David Cutler JAMA study

The Association Between Income and Life Expectancy in the United States, 2001-2014 JAMA 2016 Most cited among 20,276 articles in JAMA, Altmetric score 3,873, second only to Pres.Obama’s article.

slide-4
SLIDE 4

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Raj Chetty,...David Cutler JAMA study

The Association Between Income and Life Expectancy in the United States, 2001-2014 JAMA 2016 Most cited among 20,276 articles in JAMA, Altmetric score 3,873, second only to Pres.Obama’s article. Major Conclusions using 14 billion observations

slide-5
SLIDE 5

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Raj Chetty,...David Cutler JAMA study

The Association Between Income and Life Expectancy in the United States, 2001-2014 JAMA 2016 Most cited among 20,276 articles in JAMA, Altmetric score 3,873, second only to Pres.Obama’s article. Major Conclusions using 14 billion observations

1 Gap in life expectancy (LE) between the richest 1% and

poorest 1% of individuals was 14.6 years

slide-6
SLIDE 6

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Raj Chetty,...David Cutler JAMA study

The Association Between Income and Life Expectancy in the United States, 2001-2014 JAMA 2016 Most cited among 20,276 articles in JAMA, Altmetric score 3,873, second only to Pres.Obama’s article. Major Conclusions using 14 billion observations

1 Gap in life expectancy (LE) between the richest 1% and

poorest 1% of individuals was 14.6 years

2 Geographic differences in LE for individuals in the lowest

income quartile significantly correlated with health behaviors such as smoking (r=0.69, P.001), but not significantly correlated with access to medical care

slide-7
SLIDE 7

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Major Conclusions Contd.

3 Measures of health insurance coverage and spending (the

fraction of uninsured and risk-adjusted Medicare spending per enrollee) not significantly associated with LE for individuals in the bottom income quartile.

slide-8
SLIDE 8

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Father and Son Controversies

Are Physicians services really uncorrelated with patients health ? Health Insurance coverage does not affect life expectancy ?

  • r are these due to some statistical problems ?

This study already quoted by opponents of Obama Care!

slide-9
SLIDE 9

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

A Simple Model

The production relation is LE = β1 (M) + β2 (F) + β3 (Z) (1) Choose M and F to minimize cost of LE, i.e., the last dollars spent on a specialist and a PCP must increase LE by the same amount. Then, F ∗ = 1 + 1 r + M

∂ β1 ∂ M ∂ β2 ∂ F

(2) where LE : Expected Life Expectancy M : MDs per 10,000 population F: fraction of PCPs 1 + r : ratio of fee for services of specialists to PCPs

slide-10
SLIDE 10

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Practical Decision Rule

Rule depends on four observable variables

1 Physician density : M 2 Marginal effects of increasing physicians without changing

the proportion of PCPs: β1 (M)

3 Marginal effects of substituting PCPs by specialists:

β2 (F) and

4 Ratio of fees for services of specialists to PCPs: (1+r)

How to estimate β1 (M)&β2 (F)?

slide-11
SLIDE 11

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Big Data: Benefits and Problems

Three errors in empirical studies: (1) Sampling Errors- small samples; (2) Measurement Errors (3) Specification Errors - selection of variables and specifying functional forms. With millions and billions of observations, sampling errors will disappear; t-values will be 200 or 300 ! Many variables may be statistically significant, but are they clinically significant? No need to specify any functional form such as linear, quadratic etc. choice of functional forms could be completely data driven! No need for prior restrictions Main Problem: Sound knowledge of subject needed to measure well, assess clinical significance & use the results!

slide-12
SLIDE 12

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Data

Variables

1 unit of observation: County 2 Average Life Expectancy at age 40 for 1st and 4th income

quartiles for men and women

3 Physician Supply: Non-federal MDs/ 10 K

Population,seeing patients,75 years or less excluding DOs.

4 PCP: GP,FP,GIM,Peds., ObGyn and DOs excluding PA &

NP

5 Proportion Currently Smoking 6 Rate of self-reported exercise during the past month 7 Proportion of Medicare recipients admitted at least once

in a year

8 Number of immigrants per 10 K individuals.

Sources: AHRF and Raj et al

slide-13
SLIDE 13

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Life Expectancy, Income and Total MDs 10,000 Population - a

75.5 76.0 76.5 77.0 10 20 30 40

Male LE −1st Inc.Quart.

83.5 84.0 84.5 85.0 85.5 86.0 10 20 30 40

Male LE −4th Inc.Quart.

81.50 81.75 82.00 82.25 82.50 10 20 30 40

MDs/10,000 Population Female LE −1st Inc.Quart.

86.5 87.0 87.5 88.0 88.5 10 20 30 40

MDs/10,000 Population Female LE − 4th Inc.Quart.

slide-14
SLIDE 14

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Life Expectancy, Income and Total MDs /10,000 Population - b

74 76 78 80 15 20 25

Life.Exp.Males.Q1

80 81 82 83 84 15 20 25

Medic.IP.PerCent Life.Exp.Females.Q1

83 84 85 86 87 15 20 25

Life.Exp.Males.Q4

84 85 86 87 88 15 20 25

Medic.IP.PerCent Life.Exp.Females.Q4

slide-15
SLIDE 15

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Life Expectancy, MD Density & Specialty Distribution, Income and Sex

75 80 85 90 25 50 75 100

MD/10,000 Population Life Expectancy at Age 40 Sex

Female Male NA

PCP.percent

20 30 40 50 60 70

Income

High Low NA

slide-16
SLIDE 16

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Adjusted State Effects on Life Expectencies Referent: Massachusetts

−1 1 2 3 4 ND MN RI MT CT NY CO ME UT PA DC MD ID FL VT WA NC WI NE SD NH NJ MO VA DE MI MS CA IA SC OR TN LA IL AL OH GA KY AR AZ WY KS WV TX IN OK NM NV HI

State Decrease in LE (years) sex

Men Women

income

Poor Rich

slide-17
SLIDE 17

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Change in Life Expectancy Years (Adjusted) Physician Supply and Specialty Distribution

Massachusetts Oklahoma −0.5 0.0 0.5 1.0

10 1517 25 30 40 50

Life.Exp.Males.Q1

−0.5 0.0 0.5 1.0 10 20 30 40 50

Life.Exp.Males.Q4

Massachusetts Oklahama −1.5 −1.0 −0.5 0.0 0.5

.2 .26 .38.41 .53 .6

−0.50 −0.25 0.00 0.25 0.2 0.4 0.6 −0.5 0.0 0.5 1.0 10 20 30 40 50

Life Exp.Females.Q1

−1.0 −0.5 0.0 0.5 10 20 30 40 50

MD.10000 Life.Exp.Females.Q4

−1.5 −1.0 −0.5 0.0 0.5 0.2 0.4 0.6 −0.6 −0.3 0.0 0.3 0.2 0.4 0.6

PCP.MD.Ratio

slide-18
SLIDE 18

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Life Expectancy (Adjusted) and % Immigrants

1 2 3 4 10 20 30 40 50

Life.Exp.Males.Q1

−2 −1 10 20 30 40 50

Life.Exp.Males.Q4

1 2 3 4 10 20 30 40 50

% Immigrants Life.Exp.Females.Q1

−2.0 −1.5 −1.0 −0.5 0.0 10 20 30 40 50

% Immigrants Life.Exp.Females.Q4

Figure: 4

slide-19
SLIDE 19

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

GAM Model:Medical Care, Health Behavior And Life Expectancy

  • Dep. Var.

LE-Males(Q1) Ind.Var. edf p-value s(TotMDPop) 2.874 1.02e-07 ∗∗∗ s(PCP.MD) 2.989 1.22e-05 ∗∗∗ s(cur.smoke.q1) 2.989 7.56e-06 ∗∗∗ s(exercise.any.q1) 6.268 0.021 ∗ s(ImigPop) 5.249 < 2e-16 ∗∗∗ s(Medic.IP.%) 2.886 < 2e-16 ∗∗∗ R-sq.(adj) 0.549

  • Dev. Exp 57.4%

GCV = 1.05 Scale est. = 0.99 n = 1,359

slide-20
SLIDE 20

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Physician Density & PCP-MD Ratios

Observed Relation Policy Change 0.3 0.4 0.5 0.6 10 20 30 40

MDs per 10,000 Population PCP.MD.Ratio

slide-21
SLIDE 21

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Efficient Physician Workforce and Specialty Mix: Massachusetts and U.S.A.

Massachusetts

MD.10K.Observed =38.95 /10K MD.10K.Optimal = 37.33 /10K PCP.MD.Observed = 26.1% Specialists.Observed = 29 /10K PCP.Observed = 10 /10K Specialists.Optimal = 18 /10K PCP.Optimal = 20 /10K PCP.MD.Optimal = 51.9% 78.1 78.2 78.3 78.4 78.45 78.62 20 26.1 30 40 51.9 60 70 80

Predicted Life Expectancy(years)

U.S.

MD.10K.Observed.=24.45/10K MD.10K.Optimal = 26.48/10K PCP.MD.Observed = 30.62% Specialists.Observed = 17/10K PCP.Observed = 8 /10K Specialists.Optimal = 13/10K PCP.Optimal = 14/10K PCP.MD.Optimal = 52.56% 75.9 76.0 76.1 76.2 76.3 76.45 20 30.62 40 52.56 60 70 80

Primary Care Physicians (% of Total MDs) Predicted Life Expectancy(years)

slide-22
SLIDE 22

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Conclusions and Extensions

Conclusions

1 Life Expectancy of Poor is significantly correlated with

supply of physicians especially PCPs.

2 In almost all counties, there is shortage of PCPs and

surplus of specialists

Extensions

1 Primary care Service Area and Hospital Referral Regions 2 Include State specific policies such as Medicaid and local

government expenditures

slide-23
SLIDE 23

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Small Area Variations & Short Run Relations

Estimated Relation MD.10K=15 MD.10K=25 MD.10K=30 MD.10K=20 78.75 79.00 79.25 79.50 79.75 42 44 46 47 48 50

PCP.MD (%) Life Expectancy

Life Expectancy and PCP.MD (%) in 4 Hypothetical States

slide-24
SLIDE 24

Physician Workforce, Primary Care and Life Expectancy at 40 V.K.Chetty, Benjamin Petrie and Larry Culpepper

Our Approach

Hi Chetty, I woke up in the middle of the night thinking about what was bothering me about the papers conclusions...... Larry Hi Larry, I am glad and sorry at the same time- I disturbed your sleep but also provided some intriguing questions! This is precisely where I was quite certain that it should be the call of an expert, and not a statistician and I was looking forward to your input.