evidence for quality of surgical care in rural america samuel r. g. finlayson, md, mph
TRANSCRIPT
Overview
1. General Issues Related to Quality of Surgical Care in Rural Areas
2. Evidence for Quality of Rural Surgical Care– review of scant literature
3. Preview of new data from Mithoefer/Dartmouth collaboration
Quality in Rural Surgical Practice
• Providing high quality surgical care may not be enough to satisfy all stakeholders– Payers want evidence of high quality– Patients do, too
• Challenge is in providing evidence of high quality surgical care
Quality in Rural Surgical Practice: Challenges
• Can I just keep track of my outcomes?– Problem of Small Numbers– Practical Realities
• payer interests
• Potential Solutions– Participation in quality initiatives
• NSQIP, Michigan BC/BS, SCOAP
– Aggregate measure
Quality in Rural Surgical Practice: Challenges
• Can I just keep track of my outcomes?– Problem of Small Numbers– Practical Realities
• payer interests
• Potential Solutions– Participation in quality initiatives
• NSQIP, Michigan BC/BS, SCOAP
– Aggregate measure
Low-Volume Providers are Stuck Nearly Impossible to Demonstrate High Quality Results
When rates of adverse outcomes are low, or few procedures are performed …
… statistical power is often insufficient to show any difference between your own outcome rate and the “benchmark” rate.
The Problem of Power
Survived Died
The Nation(benchmark)
You
48,500
46 4
1500 3%
8%
difference in mortality is NOT statistically significant!
The Problem of Power
Survived Died
The Nation(benchmark)
You
48,500
46 0
1500 3%
0%
difference in mortality is NOT statistically significant!
The Problem of Power
Survived Died
The Nation(benchmark)
You
48,500
100 0
1500 3%
0%
difference in mortality is NOT statistically significant!
Quality in Rural Surgical Practice: Challenges
• Can I just keep track of my outcomes?– Problem of Small Numbers– Practical Realities
• payer interests
• Potential Solutions– Participation in quality initiatives
• NSQIP, Michigan BC/BS, SCOAP
– Aggregate measure
Practical Realities
• Payers may not care about individual results– Bariatric Surgery in Durant, OK.
• Proxies for quality are easier for insurers, align with their interests (easy approximations)– Procedure volume– Crude mortality rates– Special certifications
Quality in Rural Surgical Practice: Challenges
• Can I just keep track of my outcomes?– Problem of Small Numbers– Practical Realities
• payer interests
• Potential Solutions– Participation in quality initiatives
• NSQIP, Michigan BC/BS, SCOAP• Documents attention to quality
– Aggregate outcome measures
Aggregation to Achieve Sample Size
• Individual case volumes may be too small to demonstrate one’s results are in line with quality benchmarks
• Evaluation of surgical results in the aggregate may help – Irate letter from Iowa
Aggregating Data Tells a Story
Survived Died
Highest volume surgeon
Rest of Iowa 1331
233
4
7 3%
0.3%
HV surgeon’s mortality is higher (p<0.001)
“Are you just pissing and moaning, or can you verify what you’re saying with data?”
Evidence for Quality in Rural Surgical Practice
Mithoefer/Dartmouth
• Recall prior study of surgeon workforce based on Hospital Service Area (showed results at last symposium)– HSAs categorized along the urban-rural spectrum
using RUCA designations– Calculated surgeons per capita (age/sex-adjusted)
in each HSA– Compared surgeon workforce levels
• across specialties• across the urban-rural spectrum
6.946.54 6.30
5.52
0
2
4
6
8
10
12
Urban Large Rural Small Rural Isolated
Ge
ner
al S
urg
eon
s p
er
100
,00
0
pop
ula
tion
# of HSAs Low supply HSAs
Urban 1273
Large Rural 689
Small Rural 660
Isolated 445
Proportion of HSAs with Low Surgeon Supply (< ½ nat’l ave.)
# of HSAs Low supply HSAs
Urban 1273 254 (20%)
Large Rural 689 147 (21%)
Small Rural 660
255 (39%)
Isolated 445 295 (66%)
Proportion of HSAs with Low Surgeon Supply (< ½ nat’l ave.)
Defining Rural
• Rural-Urban Commuting Areas (RUCA)– Developed by US Dept of Agriculture, the Health
Resources and Service Administration (HRSA), and the Univ. of Washington
– Classify U.S. census tracts using measures of• population density and urbanization• size and direction of daily commuting flow
– UW developed ZIP code approximation of the RUCA codes based on an overlay of ZIP code areas on census tracts
http://www.ers.usda.gov/briefing/Rurality/RuralUrbanCommutingAreas
Hospital Service Areas (HSA)
• Developed using national Medicare inpatient claims • Defined as geographic area that includes one or
more hospitals to which local residents generally have the plurality of their inpatient admissions
• Based on zip code tabulation areas (ZCTA)– ZCTA is an approximate area representation of the US
Postal Service's ZIP code service areas created by the US Census Bureau
C
A
B
B
B
BB
B
BB
B
B
B
B
BB B
B
B
B
B
BB
B
B
B
B B
B
B
B
B
C
C
C C
C
C
C
CC
C
CC
CC
C C
C
C
C
C
C C
C C
C
C
C AA
A A
A
A
A
A
A
A
AA
A A
A
AA
A
A
A
AA
A
AA
AA B
B
B
B
B
B
B
B
C
C
C
AA
A A
A
A
AA
A
A
A
BB
BB
BB
A
A
A
A
A
A A
A
A
AA A
A
A
A 6 A’s8 B’s2 C’s
8 A’s6 B’s0 C’s
Hospital Service Areas (HSA)
• US parsed into 3067 HSAs• Reflect “health care markets”• Where people actually go for care (not
necessarily closest hospital)
Defining Rural vs. Urban HSAs
• Categorized hospital service areas into one of 4 RUCA categories– Urban– Large rural– Small rural– Isolated rural
• RUCA categorization of a given HSA determined by the plurality of that HSA’s population
Preview of New Data from Mithoefer/Dartmouth
• Effort to study outcomes– By rurality of the HSA (patient origin) – By level of surgeon supply (surg-per-capita)
• Appendectomy• A good model for studying access to surgical care
– timeline, clinical decision making– proposed as potential quality indicator
• rural patients have slightly higher rates of perforated appendectomy (adj. OR 1.11, 95%CI 1.02-1.22)*
*NIS study, to be presented at ACS Surgical Forum 2009
Mithoefer/Dartmouth
• Appendicitis Outcomes Study– National 100% sample of Medicare discharges
(age over 65, 3 years of data, >92,000 cases)– Stratified outcomes by urban-rural designation
– Specific Outcomes Studied• perforation rates• abdominal abscess rates• negative appendectomy rates
32.833.631.633.5
0
20
40
60
80
100
Pro
po
rtio
n o
f C
as
es
wit
h P
erf
ora
tio
n (
%)*
Urban/Suburban Large Town Small Town Isolated Rural
Diagnosis of Perforation with Appendectomy
*Adjusted for age, sex, race
14.4 14.7 13.112.7#
0
20
40
60
80
100
Pro
po
rtio
n o
f C
ases
wit
h A
bsc
ess
(%)*
Urban/Suburban Large Town Small Town Isolated Rural
Diagnosis of Abscess with Appendectomy
*Adjusted for age, sex, race
39.8 42.4 45.2#44.5#
0
20
40
60
80
100
Ap
pen
dec
tom
y w
ith
No
Dia
gn
osi
s o
f A
pp
end
icit
is (
%)
Urban/Suburban Large Town Small Town Isolated Rural
Negative Appendectomy Rate
Surgeon Supply and Appendicitis Outcomes
• From prior work, we know surgeons per capita for each HSA (population age sex adjusted)
• Outcomes stratified by RUCA and surgeon supply– No surgeons– ½ the national average surg-to-pop ratio– ½ to 2x national average surg-to-pop ratio– >2x national average surg-to-pop ratio
Perforation Rates and Surgeon Supply
29
30
31
32
33
34
35
36
37
none <1/2 nationalaverage
1/2 to 2xnational ave
>2x nationalaverage
overall urban/suburban small rural