Epidemiology quiz

goergesandy2023

EvaluatingHealthServices.pdf

Home >Applied Sciences homework help >Epidemiology quiz

Evaluating Health Services

David Celentano, ScD, MHS Johns Hopkins University

!  Explain how epidemiologic study designs can be used to evaluate health services and clinical interventions

!  Determine how measurement of interventions can affect how we determine effectiveness

Objectives

!  “Evaluating the Quality of Medical Care”

!  Milbank Memorial Fund Quarterly, 1965; 44(3, Part 2): 166–206

Avedis Donabedian, MD

!  Structure !  Inputs and resources into the services (organizational framework)

Structure-Process-Outcome

!  Process !  Activities themselves: referrals, admissions

Structure-Process-Outcome

!  Outcome !  Effectiveness of activities (M&M, QoL, patient satisfaction)

Structure-Process-Outcome

Donabedian’s Evaluation Model

The material in this video is subject to the copyright of the owners of the material and is being provided for educational purposes under rules of fair use for registered students in this course only. No additional copies of the copyrighted work may be made or distributed.

Evaluating Health Services

Section A

!  Efficacy !  The extent to which a specific intervention or service produces a beneficial result

under ideal conditions (based on results of an RCT)

Evaluating Health Services

!  Effectiveness !  The extent to which a specific intervention or service, as deployed in the field,

does what it is intended to for a defined population

Evaluating Health Services

!  Efficiency !  The effects or end-results achieved in relation to the effort expended—cost/benefit

(in terms of money, resources and time, a measure of economy with which a procedure of known efficacy and effectiveness is carried out)

Evaluating Health Services

!  Quantifiable

!  Relatively easy to define and diagnose

!  Population served and comparison population must be at risk for the same condition

!  Available methods of prevention and intervention must be known to “work”

Criteria for Selecting a Morbidity Index to Serve as a Measure of Effectiveness

!  Number or proportion of people immunized

!  Number or proportion of people at risk immunized

!  Number or proportion of people immunized who show serologic response

Some Possible Endpoints for Measuring Effectiveness of a Vaccine Program

!  Number or proportion of people immunized and later exposed who do not develop clinical disease

!  Number or proportion of people immunized and later exposed who do not develop clinical or subclinical disease

Some Possible Endpoints for Measuring Effectiveness of a Vaccine Program

!  Are the characteristics of the populations comparable? !  Demographically and in factors relating to risk and prognosis

!  Are the measurement methods comparable? !  Diagnostic methods and classification

Problems in Comparing the Effects of Medical Care in Two Population Groups

Classic Epidemiologic Research into Etiology

Classic Epidemiologic Research into Effectiveness

Classic Health Services Research into Effectiveness

!  Cohort 1. Randomized

Types of Epidemiologic Study Designs for Evaluating Health Services

Randomized Trial of Three Strategies of Care after Acute Stroke

Kalra et al. (2000). Lancet;356:894–9. 22

Mortality or Institutionalization at Three, Six, and Twelve Months after Stroke

Kalra et al. (2000). Lancet;356:894–9. 23

Stroke unit Stroke team Home care

Three months 15/152 (10%) 30/151 (20%) 29/146 (20%)

Six months 19/152 (13%) 37/149 (25%) 34/144 (24%)

Twelve months 21/152 (14%) 45/149 (30%) 34/144 (24%)

Kaplan-Meier Survival Curves for Different Strategies of Care after Acute Stroke

Kalra et al. (2000). Lancet;356:894–9. 24

!  Cohort 1. Randomized 2. Non-randomized

Types of Epidemiologic Study Designs for Evaluating Health Services

22-Year Cohort Study of Health Professionals

Source: Nishihara et al. (Sept 19, 2013). NEJM. 28

Outcomes Research

Section B

!  Morbidity

!  Mortality

!  Quality of life (Euro-QAL)

!  Functional status (ADL, IADL)

!  Patients’ perceptions of health status !  Symptom recognition !  Patient satisfaction and pain levels (smiley faces)

Some Health Endpoints Used in Outcomes Research

!  Hospitalization rates, re-admissions within 30 days (ACA reimbursement)

!  Outpatient (+) and emergency room visits (–)

!  Lost days of work

!  Days of restricted activity

!  Out-of-pocket expenses

Some Economic Endpoints Used in Outcomes Research

DALY like

Jollis et al., 1994, N Engl J Med;331:1625–9

In-Hospital Mortality and Rates of Bypass Surgery during Index Hospitalization According to Hospital Volume of Angioplasty Procedures Each Year

<50 procedures/ year

50–100 procedures/year

>100 procedures/ year

In-hospital mortality (%) 3.7% 3.2% 2.7%

Bypass surgery during the index hospitalization

5.3% 4.6% 3.5%

Readmission 12.7% vs. 16.8% for High- vs. Low- Volume Quartiles

Source: (2013). N Engl J Med;369:1134–42. 5

Gloves and Gowns

Antiretroviral Therapy

Studies of Disparities in Health Care

Section C

Racial Disparities in the United States

!  Asthma prevalence elevated in Puerto Ricans and Blacks

!  OR of 3 for systemic lupus erythematosis in Blacks compared to Whites

!  American Indians and Alaskan Natives twice as likely to acquire HCV as Whites

!  HIV/AIDS eight times more common in Blacks than in Whites

!  Blacks at greater risk for ESRD !  Donate organs proportionately but

more often cannot get transplants

!  Blacks account for 13% of population, but: !  70% of gonorrhea !  Half of syphilis and chlamydia

!  Why Medicare? universal access! !  Should be free of bias

NEJM, 1996;335:791

Rates of Mortality by Race, Sex, and Income among Medicare Beneficiaries 65 and Older

Source: Gornick et al. (1996). 4

Rates of Hospitalization for Ischemic Heart Disease by Race and Income among Medicare Beneficiaries 65 and Older, 1993

Source: Gornick et al. (1996). 5

Rates of Mammography by Race and Income among Medicare Beneficiaries 65 and Older, 1993

Source: Gornick et al. (1996). 6

Rates of Lower Limb Amputation by Race and Income among Medicare Beneficiaries 65 and Older, 1993

Source: Gornick et al. (1996). 7

!  Data refer to real-world populations !  The problem of external validity is minimized

!  Analysis can usually be completed quickly because the data have already been collected

!  Sample size is usually not a problem !  Except in specific subgroups

Advantages of Using Large Data Sets

!  Because data are gathered for fiscal or administrative purposes, they are often not well suited for research and may be incomplete

!  Data on independent and dependent variables may be very limited

!  Data on disease severity, details of intervention, and diagnostic coding may be inconsistent

!  Data relating to possible confounders may be inadequate or missing

Disadvantages of Large Data Sets

Evaluating Health Services Using Non- Randomized Designs

Section D

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after

Types of Epidemiologic Study Designs for Evaluating Health Services

Antibiotic Cycling in a PICU

Moss et al. (2002). Crit Care Med;30:1877–82. 3

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after •  Simultaneous comparison

Types of Epidemiologic Study Designs for Evaluating Health Services

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after •  Simultaneous comparison -  Program/no program

Types of Epidemiologic Study Designs for Evaluating Health Services

Mortality from CVD and CA by Surgical Volume

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after •  Simultaneous comparison -  Program/no program -  Utilizers/non-utilizers

Types of Epidemiologic Study Designs for Evaluating Health Services

Relationship of Neonatal Mortality to History of Prenatal Care among Adolescents in Baltimore

Prenatal care

No prenatal care

Number of births 1,397 315

Number of neonatal deaths 42 28

Neonatal deaths per 1,000 live births 30.1 88.9

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after •  Simultaneous comparison -  Program/no program -  Utilizers/non-utilizers -  Eligibles/non-eligibles

Types of Epidemiologic Study Designs for Evaluating Health Services

Design of a Study of Eligible and Non-Eligible Patients

Adjusted Prematurity Rates per 100 Live Births by Ethnicity, New York City and HIP Members

New York City HIP p

Total 6.2 5.7 0.04

White 6.0 5.5 0.06

Non-White 10.8 8.8 0.01

Perinatal Mortality by Ethnicity in New York City and HIP (Adjusted)

New York City HIP p

Total 27.9 23.1 0.01

White 27.3 22.7 0.01

Non-White 43.8 33.7 0.05

!  Cohort 1. Randomized 2. Non-randomized

•  Before/after •  Simultaneous comparison •  Combination design

Types of Epidemiologic Study Designs for Evaluating Health Services

!  After program has been started:

Combination Design

Design of a Case-Control Study for Evaluating the Effectiveness of a Vaccine

!  “Case-control study of the effectiveness of vaccination with pentavalent rotavirus vaccine in Nacaragua” !  Cases (n=300) severe rotavirus gastroenteritis cases in Nicaragua !  792 hospital controls, 851 community controls !  Vaccine coverage of RV5 reached 92% !  VE = 87% for community controls receiving three doses followed two years

•  64% in hospital controls !  VE = 85% in children <12 months at RGE onset

Pediatr Infect Dis J., 2011 Nov;30(11):e209-15

Design of a Case-Control Study for Evaluating Health Services

!  DM2 cases ages 15–34 registered in Sweden in 1983 matched with two population controls

!  1991 interviewed on retrospective data on utilization in prior three months

!  74% of cases and 19% of controls reported at least one OPD hospital visit !  OR=14 for one visit !  OR=11 for two visits !  OR=8.9 for 3+ visits

Jonsson PM et al. Diabetes and Health Service Use

20 Source: (2006). Diabet Med;13:1056–63.

One Problem in Deciding Program and Policy Priorities

!  Basic epidemiologic study designs can be used to evaluate health services

!  The focus in evaluation research is on different issues of measurement and assessment than in etiologic studies

Summary

Your feedback is very important and will be used for future revisions.

The Evaluation link is available on the lecture page.

Lecture Evaluation