Trends in the diffusion of robotic surgery in prostate, uterus, and colorectal procedures: a retrospective population-based study

Abstract

This study aimed to propose quantifiable metrics on the adoption lifecycle of robotic-assisted surgery (RAS) within and across prostate, hysterectomy, and colorectal procedures. This was a retrospective population-based cohort study of commonly performed RAS procedures in the US conducted from July 2001 to July 2015. The patients were identified from the Premier Hospital Database using International Classification of Diseases, 9th revision, Clinical Modification codes denoting prostate, uterus, and colorectal procedures. The Diffusion of Innovations theory was applied to percent RAS utilization to determine discrete eras of technology adoption. Overall and by-era patient baseline characteristics were compared between robotic and non-robotic groups. This study included a total of 2,098,440 RAS procedures comprising prostate (n = 155,342), uterus (n = 1,300,046), and colorectal (n = 643,052) procedures. Prostate (76.7%) and uterus (28.9%) procedures had the highest robotic utilization by the end of the study period and appear to be in the last adoption era (Laggard). However, robotic utilization in colorectal procedures (7.5%) was low and remained in the first era (Innovator) for a longer time (15 vs 60 vs 135 months). Whites, privately insured, patients with fewer comorbidities, and those admitted in large teaching hospitals were more likely to undergo RAS in the early study period. AS-associated patient and hospital profiles changed over time, suggesting that selected patient cohorts should be contextualized by overall adoption of a novel medical technology. The time-discretized analysis may also inform patient selection criteria and appropriate timing for clinical study stages proposed by the Idea, Development, Exploration, Assessment, Long-term study-Devices framework.

Full text

Vol.:(0123456789)
1 3
Journal of Robotic Surgery (2021) 15:275–291
https://doi.org/10.1007/s11701-020-01102-6
ORIGINAL ARTICLE
Trends inthediffusion ofrobotic surgery inprostate, uterus,
andcolorectal procedures: aretrospective population‑based study
GaryChung1· PietHinoul2· PaulCoplan1· AndrewYoo3
Received: 27 February 2020 / Accepted: 9 June 2020 / Published online: 20 June 2020
© Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract
This study aimed to propose quantifiable metrics on the adoption lifecycle of robotic-assisted surgery (RAS) within and
across prostate, hysterectomy, and colorectal procedures. This was a retrospective population-based cohort study of com-
monly performed RAS procedures in the US conducted from July 2001 to July 2015. The patients were identified from the
Premier Hospital Database using International Classification of Diseases, 9th revision, Clinical Modification codes denoting
prostate, uterus, and colorectal procedures. The Diffusion of Innovations theory was applied to percent RAS utilization to
determine discrete eras of technology adoption. Overall and by-era patient baseline characteristics were compared between
robotic and non-robotic groups. This study included a total of 2,098,440 RAS procedures comprising prostate (n = 155,342),
uterus (n = 1,300,046), and colorectal (n = 643,052) procedures. Prostate (76.7%) and uterus (28.9%) procedures had the
highest robotic utilization by the end of the study period and appear to be in the last adoption era (Laggard). However, robotic
utilization in colorectal procedures (7.5%) was low and remained in the first era (Innovator) for a longer time (15 vs 60 vs
135months). Whites, privately insured, patients with fewer comorbidities, and those admitted in large teaching hospitals
were more likely to undergo RAS in the early study period. AS-associated patient and hospital profiles changed over time,
suggesting that selected patient cohorts should be contextualized by overall adoption of a novel medical technology. The
time-discretized analysis may also inform patient selection criteria and appropriate timing for clinical study stages proposed
by the Idea, Development, Exploration, Assessment, Long-term study-Devices framework.
Keywords Robotic-assisted surgery· Prostate· Uterus· Colorectal· Diffusion of innovation· Premier Hospital database
Introduction
The process of planned and spontaneous spread of innova-
tions over time among members of a particular social system
is called diffusion [1]. The process of diffusion of innovation
[1] has been validated across surgical procedures, the use
of antibiotics, and anesthesia during surgeries [2, 3]. The
translation of novel surgical devices from the laboratory to
the operating room is vital for the progression of surgical
practice [4]. Devices that are manufactured in partnership
with clinicians and industry are expected to result in a suc-
cessful first-in-human study and attain regulatory approval,
respectively [5, 6]. However, the adoption of these devices
remains complex and poorly understood [7].
The introduction of surgical robots has provided benefits
to surgical outcomes, especially in urology wherein the high-
est proportion of RAS has been utilized in urology (34.1%),
followed by gynecology (11.0%), and endocrine surgery
(9.4%). Previous studies have shown that compared to an
open retropubic approach, robot-assisted radical prostatec-
tomy had a shorter operating time and reduced length of stay
and estimated total blood loss [8–10]. The rapid diffusion of
robotic-assisted surgery (RAS) over laparoscopy and open
approaches could be attributed to the associated benefits of
RAS including improved ergonomics, improved accessibil-
ity, visualization, and direct-to-consumer advertising by
urologist and companies [11, 12]. However, rapid technol-
ogy adoption in surgical procedures makes generating clini-
cal relevant safety and effectiveness data very challenging.
* Andrew Yoo
AYoo@ITS.JNJ.com
1 Johnson & Johnson, Medical Devices Epidemiology
andReal-World Sciences, NewBrunswick, NJ, USA
2 Ethicon, Inc., Clinical andMedical Affairs, Somerville, NJ,
USA
3 Johnson & Johnson, C-SATS, Outcomes Research
andMedical Affairs, Seattle, UnitedStates

276 Journal of Robotic Surgery (2021) 15:275–291
1 3
Despite the gradual adoption of RAS for various proce-
dures, the pattern of diffusion and extent of adoption vary by
procedure and speciality [13]. The reason could be a lack of
appropriate and accessible measures of innovation resulting
in a failure to capture the diffusion of techniques across all
healthcare disciplines [14, 15]. Moreover, the availability of
data from registries and clinical trials supporting the rapid
adoption of RAS is limited when compared to the actual
usage of these novel surgical techniques [1, 14].
Increased use of RAS in prostatectomy, uterus, and colo-
rectal procedures has been elucidated in recent studies [16,
17]. Previous studies have identified disparities in patients
undergoing RAS with respect to their demographic charac-
teristics such as race, socioeconomic status, insurance, hos-
pital types, and geographical location. [18–20] In addition,
the distribution of these characteristics change over time
and the differences tend to decrease over the study period
[13]. However, the evolution of patient populations over time
receiving RAS has not been well characterized. Moreover,
RAS adoption in prostate and uterus procedures is consid-
ered at or near saturation, which calls for a real-world evi-
dence study [16, 21]. Despite the increase in awareness of
the innovation theory and its application in healthcare [22,
23], a robust method for quantitative longitudinal analysis
of RAS diffusion is not available.
Therefore, it is imperative to assess trends, utilization
characteristics, and factors associated with adoption patterns
over time. The aim of this retrospective real-world evidence
study is to propose quantifiable metrics on the adoption
lifecycle of RAS within and across three different surgical
anatomical areas including the prostate, hysterectomy, and
colorectal procedures.
Methods
Study design andpopulation
This was a retrospective population-based cohort study of
the most commonly performed RAS procedures (i.e. pros-
tate, uterus, and colorectal procedures) in the US. The study
was conducted between July 2001 and July 2015 to capture
the primary period of robotic diffusion in the US. The study
utilized the Premier Hospital Database (Premier Inc., Char-
lotte, NC, US), which captures ~ 20% of annual US inpatient
discharges and has been described as a representative sample
of inpatient admissions.
Data collection andidentification
Individuals aged > 18years who had undergone either pros-
tate, uterus, or colorectal surgery as their primary proce-
dure were included in the study. The patients were identified
using International Classification of Diseases, 9th revision,
Clinical Modification (ICD-9-CM) codes denoting prostate,
uterus, and colorectal procedures. ICD-9-CM procedures
codes 17.4 × (17.41, 17.42, 17.43, 17.44, 17.45, and 17.49)
were used to identify robotic-assisted cases. The compre-
hensive billing data for each patient in the Premier Hospital
Database were used to separate out non-robotic and robotic-
assisted procedures.
The relevant procedures containing robotic technology
of interest were identified through an iterative technique.
First, all admissions with the robotic ICD-9-CM secondary
procedure code 17.4× were identified. Second, text mining
was used on the Premier Charge File for robotic surgery text
identifiers (e.g. Intuitive Surgical or da Vinci technology
text) and the primary procedure codes were added to the
relevant procedure set. This search algorithm allowed for
the identification either through the 17.4 ICD-9-CM proce-
dure code or through Premier Charge File text mining. After
identification of the surgical procedures codes, they were
grouped according to clinical procedural relevance. The
open surgery approach ICD-9-CM procedure codes were
identified and again grouped according to procedural rel-
evance. Cases with missing data on ethnicity, primary payer
type, hospital information, and mortality were excluded
from the analysis.
Case count adjustments
Case count adjustments were adopted to better approximate
actual national case volume for the derivation of diffusion
eras. These adjustments were applied to case count and era
calculations only. Provider-adjusted case count adjustment
was adopted for the changing number of providers enrolled
in Premier over time. An adjustment factor was calculated by
time to equalize the number of providers contributing case
counts at any given time used in the calculation of overall
and robotic case counts.
Pre-ICD-9 adjusted case counts were adopted to com-
pensate for the non-existence of ICD-9 code 17.4 prior to
Q42008. A linear regression-based adjustment factor was
calculated using the ratio of RAS procedures identified via
the chargemaster file to procedures identified via ICD-9 after
Q42008. This adjustment was applied predominantly to the
prostate procedures that occurred prior to Q42008.
Premier inpatient projection case count adjustment was
adopted by calculating an adjustment factor by time to
equalize the number of providers contributing outpatient
case counts at any given time. A sensitivity analysis was
performed utilizing these weights for all inpatient and out-
patient records to inform the calculation of case counts and
era determination.
Remaining descriptive statistics were based on unadjusted
cases. Case count adjustments were made to offset the effects

277Journal of Robotic Surgery (2021) 15:275–291
1 3
of the Centers for Medicare and Medicaid Services (CMS)
implementation of the robotic 17.4 ICD9 code in late 2008
and the effects of changing hospital enrollment volumes over
time in Premier [24].
Overall and by-era baseline characteristics of the patients,
including age, gender, race, BMI, marital, and insurance
status (Medicare, Medicaid, Private Insurance, Other) were
collected. To account for baseline health status, Elixhauser
(ELIX) and Charlson Comorbidity Score (CCS) were
recorded. In addition, the relevant hospital characteristics
including the number of beds, type, location, and region
were also included.
Determination ofera
Rogers diffusion process was used to assess RAS adoption
using discrete-time periods [1]. The time periods, accord-
ing to the Law of Diffusion of Innovation, are the innovator
(I), early adopter (EA), early majority (EM), late majority
(LM), and laggard (L) phases. By definition, the I, EA, and
EM time periods correspond to when 2.5, 16, and 50% of the
population have adopted a novel technology, respectively.
LM and L phases of the RAS adoption were also evaluated.
The RAS utilization, defined as the percent of all pro-
cedures with identified robotic technology, across the
five major eras were determined by a combined quantita-
tive–qualitative approach. Initial cutoffs were estimated
using the minima/maxima of the first and second deriva-
tives of the smoothed percent utilization curves, effectively
corresponding to curve inflection points. Smoothing percent
utilization using a rolling mean was necessary to reduce the
effect of transitory micro-trends on the estimation of the
macro-inflection points.
The Idea, Development, Exploration, Assessment, Long-
term study-Devices (IDEAL-D) Framework and Recom-
mendations, a model for integrated stepwise evaluation of
maturing interventions, was used to evaluate the surgical
procedures. The IDEAL Framework and Recommendations
were designed to provide an evaluation pathway for generat-
ing and analyzing data throughout the life cycle of surgical
innovations [25–28].
Statistical analysis
A descriptive summary of the patient and hospital character-
istics was performed for the patients (either prostate, uterus,
and colorectal) undergoing RAS. Categorical variables were
summarized by counts and the percentage of patients within
each category of a variable. Continuous variables were sum-
marized by the mean and standard deviation of the vari-
able distribution. Medians and other percentile information
were reported for the variables which were not normally
distributed. Odds ratios (ORs) and risk ratios (RRs) were
calculated from the descriptive statistics to provide esti-
mated measurements of effect. P values were not reported
for these data, as information from these variables is descrip-
tive in nature. Any statistical tests not otherwise specified
were two-sided at an α level of 0.05. All analyses were per-
formed using R.
Results
This study included a total of 2,098,440 robotic-assisted
procedures, comprising prostate (n = 155,342) uterus
(n = 1,300,046) and colorectal (n = 643,052) performed
between July 2001 and July 2015. The demographic, clini-
cal, and provider characteristics of the study population and
hospital characteristics are described in Table1.
Out of 155,342 prostate procedures, 79,227 (51.0%) had
robotic indicators, with discharges ranging from 2002Q2
to 2015Q2. The mean age of the patients in the robotic
prostate group was 61.6 ± 7.18years compared to the non-
robotic group (63.7 ± 8.45years). Of the 1,300,046 uterus
procedures, 158,355 (12.2%) had robotic indicators, with
discharges ranging from 2001Q3 to 2015Q2. Major differ-
ences for robotic uterus included higher age (47.9 ± 11.9 vs
46.2 ± 11.6years) compared to the non-robotic group. Over-
all, colorectal had 643,052 procedures, of which, 11,500
(1.79%) had robotic indicators, with discharges ranging from
2001Q4 to 2015Q2 (Tables2, 3, 4).
Discharge volumes bycalendar quarter
Prostate robotic procedures
Prostate robotic procedures grew significantly in the mid-
2000s and started to overtake non-robotic procedures in the
late-2000s. A general downtrend was observed in the overall
volume of prostate surgeries. Sensitivity analysis using Pre-
mier inpatient adjustment did not substantially change the
contour of these curves; however, overall discharge volumes
increased substantially. For example, the overall 2010Q1
volume increased from ~ 2700 discharges to ~ 18,000 dis-
charges (Fig.1a).
Uterus robotic procedures
Adoption of robotics for the uterus procedures started in
small volumes around 2005 and quickly accelerated in the
late 2000s, around the time when robotic prostate volume
growth started slowing down. A general downtrend was
observed in uterus surgeries. Similar to the prostate, Pre-
mier inpatient adjustment did not substantially change curve
contours; however, substantially increased overall discharge

278 Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 1 Demographic, clinical, provider, and hospital characteristics of robotic vs non-robotic procedures
Characteristics Prostate Uterus Colorectal
Overall
(n = 155,342)
Robotic
(n = 79,227)
Non-robotic
(n = 76,115)
Overall
(n = 1,300,046)
Robotic
(n = 158,335)
Non-robotic
(n = 1,141,691)
Overall
(n = 643,052)
Robotic
(n = 11,500)
Non-robotic
(n = 631,552)
Age (years;
mean ± SD)
62.6 ± 7.9 61.6 ± 7.18 63.7 ± 8.45 46.4 ± 11.6 47.9 ± 11.9 46.2 ± 11.6 62.5 ± 16.7 60.8 ± 14.4 62.5 ± 16.8
Gender (%)
Male 100 100 100 (0) 0 (0) 0 (0) 0 (46.7) 46.7 (47.1) 47.1 (46.7) 46.7
Marital status (%)
Married (70.8) 70.7 (71.4) 71.3 (70.1) 70 (56.5) 56.3 (56.1) 56 (56.5) 56.4 (49.1) 49.1 (55.4) 55.3 (49) 49
Other (12) 12 (11.3) 11.3 (12.7) 12.7 (11.2) 11.2 (10.6) 10.6 (11.3) 11.3 (11.4) 11.4 (7.21) 7.2 (11.4) 11.4
Single (17.2) 17.2 (17.2) 17.2 (17.2) 17.2 (32.3) 32.2 (33.3) 33.2 (32.2) 32.1 (39.4) 39.4 (37.4) 37.3 (39.5) 39.5
Race (%)
Black (11.9) 11.9 (11.2) 11.2 (12.6) 12.6 (15.2) 15.2 (13.3) 13.3 (15.5) 15.5 (9.5) 9.5 (8.02) 8 (9.5) 9.5
Hispanic (1.8) 1.8 (0.902) 0.9 (2.6) 2.6 (3.5) 3.5 (1.2) 1.2 (3.8) 3.8 (2.7) 2.7 (1) 1 (2.8) 2.8
Other (15.7) 15.7 (14.7) 14.7 (16.8) 16.8 (17.2) 17.2 (15.5) 15.5 (17.5) 17.5 (15.9) 15.9 (12.1) 12.1 (15.9) 15.9
White (70.6) 70.5 (73.1) 73 (68) 68 (64.1) 64.1 (70) 69.9 (63.2) 63.2 (71.9) 71.9 (78.9) 78.7 (71.8) 71.8
BMI (%)
≥ 70% (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0
Between 0
and 29.9
(23.7) 0.9 (23.2) 1.3 (28.6) 0.6 (8.89) 0.4 (5.94) 0.6 (10.8) 0.4 (37.8) 2.8 (23.1) 3 (37.8) 2.8
Between 30
and 34.9
(39.5) 1.5 (41.1) 2.3 (33.3) 0.7 (20) 0.9 (19.8) 2 (21.6) 0.8 (21.6) 1.6 (28.5) 3.7 (21.6) 1.6
Between 35
and 39.9
(23.7) 0.9 (23.2) 1.3 (19) 0.4 (24.4) 1.1 (23.8) 2.4 (24.3) 0.9 (17.6) 1.3 (22.3) 2.9 (17.6) 1.3
Between 40
and 44.9
(10.5) 0.4 (10.7) 0.6 (14.3) 0.3 (28.9) 1.3 (25.7) 2.6 (29.7) 1.1 (16.2) 1.2 (16.9) 2.2 (16.2) 1.2
Between 45
and 49.9
(2.63) 0.1 (1.79) 0.1 (4.76) 0.1 (8.89) 0.4 (12.9) 1.3 (8.11) 0.3 (4.05) 0.3 (6.15) 0.8 (4.05) 0.3
Between 50
and 59.9
(0) 0 (0) 0 (0) 0 (6.67) 0.3 (9.9) 1 (5.41) 0.2 (2.7) 0.2 (3.08) 0.4 (2.7) 0.2
Between 60
and 69.9
(0) 0 (0) 0 (0) 0 (2.22) 0.1 (1.98) 0.2 (0) 0 (0) 0 (0) 0 (0) 0
Unknown 96.1 94.4 98 95.5 89.9 96.2 92.4 86.9 92.5
Payer (%)
Commercial 55.8 60.6 50.9 71.1 70.7 71.1 37.1 45.7 36.9
Medicaid 1.7 1.7 1.8 9.2 8.7 9.3 6.2 6.4 6.2
Medicare 37.4 32.4 42.6 11 13.2 10.7 49.8 42.6 49.9
Other 5 5.4 4.7 8.7 7.4 8.9 7 5.3 7
APR-DRG mortality (%)
Extreme (0.445) 0.4 (0.319) 0.3 (0.583) 0.5 (0.434) 0.3 (0.287) 0.1 (0.407) 0.3 (10.9) 10.8 (3.52) 3.5 (11) 10.9
Major (1.45) 1.3 (1.06) 1 (1.98) 1.7 (1.3) 0.9 (1.43) 0.5 (1.22) 0.9 (16.2) 16 (7.54) 7.5 (16.3) 16.1
Minor (89.7) 80.6 (92.2) 86.7 (86.7) 74.3 (93.2) 64.5 (91.1) 31.8 (93.5) 69 (46.1) 45.6 (65.5) 65.2 (45.7) 45.2
Moderate (8.45) 7.6 (6.38) 6 (10.7) 9.2 (5.06) 3.5 (7.16) 2.5 (4.88) 3.6 (26.9) 26.6 (23.4) 23.3 (27) 26.7
Unknown 10 6.1 14.2 30.9 65.1 26.2 1.1 0.6 1.1
MS-DRG CC/MCC (%)
Outpatient (10) 10 (6.09) 6.1 (14.2) 14.2 (30.9) 30.9 (65.1) 65.1 (26.2) 26.2 (1.1) 1.1 (0.601) 0.6 (1.1) 1.1
With (18.8) 18.8 (14.5) 14.5 (23.3) 23.3 (14) 14 (8.6) 8.6 (14.8) 14.8 (70.3) 70.1 (53.9) 53.8 (70.6) 70.4
Without (71.2) 71.1 (79.4) 79.5 (62.5) 62.5 (55.1) 55 (26.3) 26.3 (59) 59 (28.6) 28.5 (45.5) 45.5 (28.3) 28.2
ELIX
(mean ± SD)
2 ± 1.09 2.06 ± 1.05 1.94 ± 1.13 0.874 ± 1.17 1.12 ± 1.32 0.841 ± 1.15 2.39 ± 1.96 2.12 ± 1.7 2.4 ± 1.96
CCS
(mean ± SD)
2.29 ± 0.992 2.36 ± 0.835 2.22 ± 1.13 0.432 ± 1.09 0.628 ± 1.21 0.405 ± 1.07 2.1 ± 2.38 1.87 ± 2.1 2.1 ± 2.38
Region (%)
East North
Central
9.8 9.6 9.9 12 14.4 11.7 12.1 9.5 12.1
East-South-
Central
8.7 11.4 5.9 7.5 10.2 7.1 6.2 4.7 6.2
Middle-
Atlantic
14.5 19.1 9.8 9.5 7 9.8 14.2 18.2 14.2

279Journal of Robotic Surgery (2021) 15:275–291
1 3
volumes. For example, the overall 2012Q1 volume increased
from ~ 22,000 discharges to ~ 100,000 discharges (Fig.1b).
Colorectal robotic procedures
Overall robotic volumes were still low as of 2015 relative
to all colorectal procedures. Relative utilizations by specific
procedures were still low, with only rectal excisions and
repairs exceeding an overall utilization rate of 30%. Similar
to the prostate, Premier inpatient adjustment did not sub-
stantially change curve contours but substantially increased
overall discharge volumes. For example, the overall 2014Q1
volume increased from ~ 11,000 discharges to ~ 77,000 dis-
charges (Fig.1c).
Adoption ofRAS inprostate procedures
A progressive increase in the adoption of robotic assistance
was observed for prostate surgeries in the US from 2001 to
2015. According to the Law of Diffusion, RAS in prostate
spanned all five adoption eras as follows: era-I from 2002Q2
to 2003Q2 (15months, n = 63, 0.667% RAS utilization), era-
EA from 2003Q3 to 2005Q2 (24months, n = 1275, 8.27%),
era-EM from 2005Q3 to 2007Q3 (27months, n = 5376,
22.8%), era-LM from 2007Q4 to 2011Q3 (48months,
n = 32,219, 58.7%), and era-L from 2011Q4 to 2015Q2
(45months, n = 40,294, 77.4%). Unique to prostate proce-
dures, RAS utilization, as of 2015Q2, was 76.7%. (Table2;
Fig.2a).
The smallest mean age difference between robotic
and non-robotic procedures was observed in the EA
era (− 1.8 years), followed by EM (−2.3 years), LM
(−3.1years), I (−3.6years), and L (−4.3years) era. In
addition, RAS in era-I was 3.76 times more likely to occur in
Whites compared to Blacks and Hispanics (OR 3.76) com-
pared to later eras (EA 1.18, EM 1.41, LM 1.42, L 1.30).
More than 90% of the discharges had commercial and
Medicare payers across all eras. However, the commercial
was the preferred payer for RAS across all eras. RAS era-I
strongly biased to commercial (OR 4.42) while the transition
to era-EA coincided with a large shift toward Medicare (OR
1.56). Subsequent eras gradually saw the rebalancing toward
commercial payers (EM 1.72, LM 1.97, and L 2.38) with the
growing volume and, apparently, expanding patient profiles.
Relatively lower RAS inpatient complications were observed
in era-I (OR 0.493), elevated after the transition to era-EA
(OR 0.561), then gradually declined as the eras progressed
(EM 0.605, LM 0.463, L 0.362).
For hospital demographics, there was a significant decline
in teaching hospitals after exiting era-I (RR of teaching in
Table 1 (continued)
Characteristics Prostate Uterus Colorectal
Overall
(n = 155,342)
Robotic
(n = 79,227)
Non-robotic
(n = 76,115)
Overall
(n = 1,300,046)
Robotic
(n = 158,335)
Non-robotic
(n = 1,141,691)
Overall
(n = 643,052)
Robotic
(n = 11,500)
Non-robotic
(n = 631,552)
Mountain 5.2 3.7 6.8 4.7 5.3 4.7 4.4 5.2 4.4
New England 3.4 4.7 2.1 2.9 4 2.7 3.2 3.3 3.2
Pacific 12.5 11.2 13.9 12.7 10 13.1 12.8 6.7 12.9
South-
Atlantic
31.8 29.3 34.3 33.4 30.3 33.8 31.1 40.8 30.9
West North
Central
5.7 4.8 6.6 6.1 5.2 6.2 6 2.9 6
West South
Central
8.4 6.1 10.8 11.2 13.6 10.9 10 8.6 10.1
Hospital size (%)
000–099 1.1 0.3 1.9 5 0.9 5.6 3.8 1.8 3.8
100–199 7.6 5.4 10 11 8.3 11.4 11 6.5 11.1
200–299 15.7 15.7 15.6 15.6 16.6 15.4 16.3 16.3 16.3
300–399 19.9 16.7 23.2 22 23.8 21.7 20.9 18.5 21
400–499 16.3 16.8 15.8 15.5 14.7 15.6 15.2 12.5 15.3
500+ 39.4 45.1 33.5 31 35.7 30.3 32.8 44.5 32.6
Teaching hospital (%)
Yes 44.8 50.3 39.1 38.7 41.2 38.3 41.8 46.5 41.7
Urban
Yes 94 97 90.8 89.9 95 89.2 90.1 95.4 90
Format of (##.#)##.# represents (% excluding missing)%
APR-DRG all patient refined diagnosis related group, BMI body mass index, CCS Charlson Comorbidity Score, ELIX Elixhauser, MS-DRG CC/
MCC medicare severity diagnosis related groups complication or comorbidity (CC) or a major complication or comorbidity (MCC), SD standard
deviation

280 Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 2 Demographic, clinical, provider, and hospital characteristics of prostate procedure by era
Characteristic Robotic Non-robotic
Innovator (n = 63) Early adopter
(n = 1275)
Early majority
(n = 5376)
Late majority
(n = 32,219)
Laggard
(n = 40,294)
Innovator
(n = 9383)
Early adopter
(n = 14,140)
Early majority
(n = 18,161)
Late majority
(n = 22,690)
Laggard (n = 11,741)
Era 2002Q2–2003Q2 2003Q3–2005Q2 2005Q3–2007Q3 2007Q4–2011Q3 2011Q4–2015Q2 2002Q1–2003Q2 2003Q3–2005Q2 2005Q3–2007Q3 2007Q4–2011Q3 2011Q4–2015Q2
Age (years;
mean ± SD)
58.9 ± 6.7 60.5 ± 7.42 60.6 ± 7.05 61.3 ± 7.12 61.9 ± 7.22 62.5 ± 7.91 62.3 ± 8.03 62.9 ± 8.22 64.4 ± 8.55 66.2 ± 8.83
Marital status (%)
Married (82.5) 82.5 (77.3) 77.3 (69.1) 69.1 (74.8) 74.7 (68.9) 68.8 (75) 75.1 (73.4) 73.4 (71.5) 71.4 (68.6) 68.5 (62.5) 62.3
Other (0) 0 (5.8) 5.8 (15.6) 15.6 (8.01) 8 (13.6) 13.6 (10.3) 10.3 (11.2) 11.2 (12.9) 12.9 (12.6) 12.6 (16.5) 16.5
Single (17.5) 17.5 (16.9) 16.9 (15.3) 15.3 (17.2) 17.2 (17.5) 17.5 (14.7) 14.7 (15.4) 15.4 (15.6) 15.6 (18.8) 18.8 (21) 20.9
Missing <0.001 <0.001 <0.001 0.1 0.2 <0.001 0.1 <0.001 0.1 0.2
Race (%)
Black (4.8) 4.8 (11.3) 11.3 (8.9) 8.9 (10.7) 10.7 (12) 12 (10.4) 10.4 (11.7) 11.7 (12) 12 (13.3) 13.3 (14.8) 14.8
Hispanic (0) 0 (3.1) 3.1 (1.5) 1.5 (1.9) 1.9 (0) 0 (4.9) 4.9 (2.9) 2.9 (2.5) 2.5 (3) 3 (0) 0
Other (12.7) 12.7 (5.7) 5.7 (21.9) 21.9 (14.4) 14.4 (14.3) 14.3 (14.8) 14.8 (16.7) 16.7 (18.4) 18.4 (17.2) 17.2 (15.2) 15.2
White (82.5) 82.5 (79.9) 79.9 (67.7) 67.7 (73) 73 (73.6) 73.5 (69.9) 69.8 (68.7) 68.7 (67.1) 67.1 (66.5) 66.5 (69.9) 69.8
Unknown <0.001 <0.001 <0.001 <0.001 0.2 <0.001 <0.001 <0.001 <0.001 0.1
Payer (%)
Commercial 85.7 69.6 67.6 62.6 57.7 57.2 57.7 55.2 47.4 37.8
Medicaid <0.001 0.5 0.9 1.3 2.1 1.2 1.3 1.5 1.9 3.3
Medicare 12.7 28.1 27.6 30.6 34.7 37.5 36.4 38.7 45.7 54.1
Other 1.6 1.8 4 5.5 5.5 4.1 4.6 4.6 5 4.8
APR-DRG mortality (%)
Extreme (0) 0 (0.5) 0.5 (0.209) 0.2 (0.427) 0.4 (0.32) 0.3 (0.31) 0.3 (0.316) 0.3 (0.442) 0.4 (0.881) 0.7 (1.27) 0.9
Major (1.6) 1.6 (0.6) 0.6 (0.94) 0.9 (1.07) 1 (1.07) 1 (1.03) 1 (0.948) 0.9 (1.44) 1.3 (2.77) 2.2 (4.23) 3
Minor (92.1) 92.1 (94) 94 (93.1) 89.1 (92.2) 86.3 (92.1) 86.4 (89.1) 86.2 (89.5) 84.9 (87.7) 79.4 (85.4) 67.9 (80) 56.7
Moderate (6.3) 6.3 (4.9) 4.9 (5.75) 5.5 (6.3) 5.9 (6.5) 6.1 (9.51) 9.2 (9.27) 8.8 (10.4) 9.4 (10.9) 8.7 (14.5) 10.3
Unknown <0.001 <0.001 4.3 6.4 6.2 3.3 5.1 9.5 20.5 29
MS-DRG CC/MCC (%)
Outpatient (0) 0 (0) 0 (4.3) 4.3 (6.41) 6.4 (6.2) 6.2 (3.3) 3.3 (5.1) 5.1 (9.5) 9.5 (20.5) 20.5 (29) 29
With (15.9) 15.9 (15.8) 15.8 (15.8) 15.8 (14.3) 14.3 (14.4) 14.4 (26.8) 26.8 (23.8) 23.8 (22.3) 22.3 (22.3) 22.3 (23.7) 23.7
Without (84.1) 84.1 (84.2) 84.2 (79.9) 79.9 (79.3) 79.2 (79.4) 79.4 (69.9) 69.9 (71.1) 71.1 (68.2) 68.2 (57.2) 57.2 (47.3) 47.3
NA
ELIX (mean ± SD) 1.76 ± 0.946 1.87 ± 0.926 1.93 ± 0.993 2.03 ± 1.01 2.11 ± 1.09 1.78 ± 0.974 1.9 ± 1.02 1.86 ± 1.17 2.03 ± 1.15 2.1 ± 1.26
CCS (mean ± SD) 2.27 ± 0.7 2.3 ± 0.735 2.31 ± 0.835 2.36 ± 0.803 2.38 ± 0.862 2.25 ± 1.05 2.28 ± 1.0 2.14 ± 1.17 2.26 ± 1.12 2.14 ± 1.26
Region (%)
East North
Central
<0.001 5.3 9.1 9.6 9.9 8.6 12.9 11.5 8.1 8.3
East-South-
Central
<0.001 26 10.4 9.2 12.8 5.4 5 5.7 6.9 5.7
Middle-Atlantic 96.8 45 37.6 20.1 14.9 10.3 12.4 9.2 8.8 8.8

281Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 2 (continued)
Characteristic Robotic Non-robotic
Innovator (n = 63) Early adopter
(n = 1275)
Early majority
(n = 5376)
Late majority
(n = 32,219)
Laggard
(n = 40,294)
Innovator
(n = 9383)
Early adopter
(n = 14,140)
Early majority
(n = 18,161)
Late majority
(n = 22,690)
Laggard (n = 11,741)
Mountain <0.001 <0.001 <0.001 3.8 4.3 6.2 6.2 8.1 6.5 6.6
New England <0.001 <0.001 4.2 4.1 5.3 1.9 2.6 2 1.6 2.7
Pacific <0.001 4.9 9.2 14 9.5 12 9.8 15.5 16.8 12.5
South-Atlantic 3.2 18.4 27 28.5 30.6 36.3 33 31.2 34.5 38.8
West North
Central
<0.001 0.2 1.5 5.3 5.1 6.4 7.1 6.4 6.2 7.1
West South
Central
<0.001 <0.001 0.9 5.5 7.5 12.8 11 10.4 10.7 9.5
Hospital size (%)
000–099 <0.001 <0.001 <0.001 0.3 0.3 1.9 1.6 1.7 2 2.5
100–199 <0.001 <0.001 <0.001 5.1 6.5 9.3 7.7 8.9 10.4 14.4
200–299 3.2 0.1 3.4 14.3 19 13.9 14.4 14.6 15.9 19.3
300–399 <0.001 5.8 16.5 16.5 17.3 26.8 26.4 20.6 21.8 23.4
400–499 <0.001 35.4 19.1 17.8 15.2 13.7 13.5 17 18 13.9
500+ 96.8 58.7 61 46 41.7 34.5 36.3 37.2 31.9 26.5
Teaching hospital (%)
Yes 96.8 61.3 64.8 53 45.9 41.3 45.3 41 35.5 33.8
Urban (%)
Yes 100 100 98.5 97.1 96.6 91.5 93.1 92.3 89.9 86.7
Format of (##.#)##.# represents (% excluding missing)%
APR-DRG all patient refined diagnosis related group, BMI body mass index, CCS Charlson Comorbidity Score, ELIX Elixhauser, MS-DRG CC/MCC medicare severity diagnosis related groups
complication or comorbidity (CC) or a major complication or comorbidity (MCC), SD standard deviation

282 Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 3 Demographic, clinical, provider, and hospital characteristics of uterus procedure by era
Characteristic Robotic Non-robotic
Innovator (n = 308) Early adopter
(n = 1756)
Early majority
(n = 12,546)
Late majority
(n = 49,066)
Laggard
(n = 94,679)
Innovator
(n = 393,615)
Early adopter
(n = 154,259)
Early majority
(n = 171,591)
Late majority
(n = 190,582)
Laggard
(n = 231,644)
Era 2001Q3–2006Q2 2006Q2–2008Q1 2008Q2–2010Q1 2010Q2–2012Q2 2012Q3–2015Q2 2001Q3–2006Q2 2006Q2–2008Q1 2008Q2–2010Q1 2010Q2–2012Q2 2012Q3–2015Q2
Age (years;
mean ± SD)
46.6 ± 10.4 48.9 ± 12.5 48 ± 12 47.6 ± 11.7 48 ± 11.9 45.7 ± 11.5 46.1 ± 11.4 46.4 ± 11.5 46.5 ± 11.6 47 ± 11.7
Marital status (%)
Married (59.8) 59.7 (52.4) 52.3 (61.2) 61.1 (56) 55.9 (55.5) 55.3 (59.3) 59.3 (57.5) 57.4 (57.1) 57 (54.4) 54.2 (52.4) 52.2
Other (9.41) 9.4 (16.3) 16.3 (4.11) 4.1 (11) 11 (11.2) 11.2 (10.8) 10.8 (11.8) 11.8 (9.62) 9.6 (11.5) 11.5 (13) 13
Single (30.8) 30.8 (31.3) 31.3 (34.7) 34.6 (33) 32.9 (33.3) 33.2 (29.9) 29.9 (30.7) 30.7 (33.3) 33.2 (34.1) 34 (34.6) 34.5
Missing <0.001 0.2 0.2 0.2 0.3 0.1 0.1 0.2 0.2 0.3
Race (%)
Black (5.2) 5.2 (8.51) 8.5 (11) 11 (13.1) 13.1 (13.9) 13.9 (14.6) 14.6 (14.6) 14.6 (15.8) 15.8 (15.8) 15.8 (17.2) 17.2
Hispanic (0.6) 0.6 (7.21) 7.2 (6.2) 6.2 (2) 2 (0) 0 (5.3) 5.3 (5.3) 5.3 (5.6) 5.6 (2.3) 2.3 (0) 0
Other (21.8) 21.8 (25.3) 25.3 (11) 11 (16.2) 16.2 (15.6) 15.6 (16.3) 16.3 (17.7) 17.7 (16) 16 (18.8) 18.8 (19.3) 19.3
White (72.4) 72.4 (59) 58.9 (71.8) 71.8 (68.7) 68.7 (70.5) 70.4 (63.8) 63.8 (62.4) 62.4 (62.6) 62.6 (63.1) 63.2 (63.4) 63.3
Unknown <0.001 <0.001 <0.001 <0.001 0.2 <0.001 <0.001 <0.001 <0.001 0.2
Payer (%)
Commercial 75.6 74.5 72.7 72 69.8 74.4 74.1 71.1 67.8 66.3
Medicaid 9.7 5.5 7.3 7.9 9.4 7.5 7.6 8.8 10.8 12.8
Medicare 9.1 13.5 12.9 12.5 13.6 9.7 9.9 10.7 11.5 12.3
Other 5.5 6.5 7.1 7.6 7.3 8.5 8.4 9.4 10 8.6
APR-DRG mortality (%)
Extreme (0) 0 (0.27) 0.2 (0.155) 0.1 (0.454) 0.2 (0.395) 0.1 (0.222) 0.2 (0.352) 0.3 (0.394) 0.3 (0.664) 0.4 (0.621) 0.3
Major (0.316) 0.3 (1.08) 0.8 (0.929) 0.6 (1.36) 0.6 (1.98) 0.5 (0.889) 0.8 (1.17) 1 (1.45) 1.1 (1.83) 1.1 (2.07) 1
Minor (95.9) 90.9 (90.3) 67 (92.9) 60 (91.8) 40.5 (89.7) 22.7 (94.6) 85.1 (93.9) 80.1 (93.3) 71 (92) 55.4 (91.1) 44
Moderate (3.8) 3.6 (8.36) 6.2 (6.04) 3.9 (6.35) 2.8 (7.91) 2 (4.33) 3.9 (4.57) 3.9 (4.86) 3.7 (5.48) 3.3 (6.21) 3
Unknown 5.2 25.7 35.4 56 74.7 10.1 14.7 23.9 39.8 51.7
MS-DRG CC/MCC (%)
Outpatient (5.2) 5.2 (25.7) 25.7 (35.3) 35.3 (56) 56 (74.7) 74.7 (10.1) 10.1 (14.7) 14.7 (23.9) 23.9 (39.8) 39.8 (51.7) 51.7
With (9.7) 9.7 (14.5) 14.5 (12.8) 12.8 (10.1) 10.1 (7.2) 7.2 (15.4) 15.4 (15.3) 15.3 (15.5) 15.5 (14) 14 (13.3) 13.3
Without (85.1) 85.1 (59.8) 59.8 (51.9) 51.9 (33.9) 33.9 (18.1) 18.1 (74.5) 74.5 (70) 69.9 (60.6) 60.5 (46.1) 46.1 (35) 35
ELIX
(mean ± SD)
0.831 ± 1.1 1.11 ± 1.33 1.09 ± 1.29 1.08 ± 1.3 1.14 ± 1.33 0.714 ± 1.05 0.815 ± 1.15 0.896 ± 1.17 0.911 ± 1.18 0.972 ± 1.23
CCS (mean ± SD) 0.422 ± 0.86 0.779 ± 1.36 0.673 ± 1.26 0.609 ± 1.2 0.629 ± 1.21 0.375 ± 1.07 0.394 ± 1.06 0.428 ± 1.09 0.417 ± 1.07 0.436 ± 1.07
Region (%)
East North
Central
6.5 15.2 18.1 14.8 13.7 11.6 12.3 11 11.9 11.9
East-South-
Central
5.2 8.5 6.1 9 11.4 7.5 6.4 6.2 5.7 8.9
Middle-Atlantic 16.2 19.6 9.3 7 6.4 9.3 10 10.9 9.9 9.9

283Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 3 (continued)
Characteristic Robotic Non-robotic
Innovator (n = 308) Early adopter
(n = 1756)
Early majority
(n = 12,546)
Late majority
(n = 49,066)
Laggard
(n = 94,679)
Innovator
(n = 393,615)
Early adopter
(n = 154,259)
Early majority
(n = 171,591)
Late majority
(n = 190,582)
Laggard
(n = 231,644)
Mountain <0.001 <0.001 2.1 4.3 6.4 3.3 5.4 6 5.4 4.9
New England 0.3 0.3 1.8 4.6 4.1 1.8 2 2.5 3.7 4
Pacific 64.3 20.4 12.9 11.4 8.5 11.7 14.5 14.3 14.6 12.3
South-Atlantic 7.5 31.7 32 30.8 29.9 36.4 31.7 32.3 32.6 32.8
West North
Central
<0.001 1.3 6.1 5.8 4.9 6.6 6.6 6.5 6.1 5.2
West South
Central
<0.001 3.1 11.6 12.4 14.7 11.8 11.2 10.2 10.1 10.1
Hospital size (%)
000–099 1.6 0.5 <0.001 0.7 1.2 5.6 5.4 5.4 5.6 6
100–199 <0.001 0.2 3.9 7.4 9.4 9.6 9.5 10.5 13.5 14.8
200–299 43.8 10.4 14.4 15.3 17.6 13.9 13.9 15.3 16.9 17.7
300–399 8.1 12 16.8 22.4 25.7 23.5 22.7 21.5 20.3 19.2
400–499 25.6 19.4 14.8 15.5 14.1 15.7 17.1 17.3 15.3 13.2
500+ 20.8 57.6 50 38.6 31.9 31.7 31.4 30.2 28.4 28.9
Teaching hospital
Yes 28.2 57.3 50 45.3 37.6 39 37.1 37.9 38.6 38.2
Urban
Yes 98.4 98.2 95 95.3 94.8 89.1 90.4 90.6 89 87.9
Format of (##.#)##.# represents (% excluding missing)%
APR-DRG all patient refined diagnosis related group, BMI body mass index, CCS Charlson Comorbidity Score, ELIX Elixhauser, MS-DRG CC/MCC medicare severity diagnosis related groups
complication or comorbidity (CC) or a major complication or comorbidity (MCC), SD standard deviation

284 Journal of Robotic Surgery (2021) 15:275–291
1 3
Table 4 Demographic, clinical, provider, and hospital characteristics of the colorectal procedure by era
Characteristic Robotic Non-robotic
Innovator (n = 2428) Early adopter (n = 9072) Innovator (n = 440,345) Early Adopter (n = 191,207)
Era (range) 2001Q4–2011Q4 2012Q1–2015Q2 2001Q4–2011Q4 2012Q1–2015Q2
Age (years; mean ± SD) 60.9 ± 16.4 60.8 ± 13.9 62.7 ± 16.9 62.2 ± 16.4
Gender
Male (44.9) 44.9 (47.6) 47.6 (46.8) 46.8 (46.5) 46.5
Marital status (%)
Married (55.2) 55.2 (55.5) 55.3 (49.6) 49.6 (47.5) 47.4
Other (5.2) 5.2 (7.82) 7.8 (11.4) 11.4 (11.5) 11.5
Single (39.6) 39.6 (36.7) 36.6 (38.9) 38.9 (40.9) 40.8
Race (%)
Black (6.41) 6.4 (8.43) 8.4 (9.41) 9.4 (9.71) 9.7
Hispanic (4.5) 4.5 (0) 0 (3.9) 3.9 (0) 0
Other (10.4) 10.4 (12.6) 12.6 (16.4) 16.4 (14.8) 14.8
White (78.7) 78.6 (78.9) 78.7 (70.3) 70.2 (75.5) 75.4
Payer (%)
Commercial 41.7 46.8 37.6 35.6
Medicaid 7.1 6.2 5.5 7.8
Medicare 45.6 41.8 50 49.7
Other 5.6 5.2 7 6.9
APR-DRG mortality (%)
Extreme (7.12) 7.1 (2.52) 2.5 (10.7) 10.6 (11.9) 11.7
Major (10.8) 10.8 (6.64) 6.6 (16.1) 15.9 (16.7) 16.5
Minor (57.2) 57 (67.8) 67.4 (45.4) 44.9 (46.6) 46
Moderate (24.9) 24.8 (23) 22.9 (27.9) 27.6 (24.8) 24.5
Unknown 0.2 0.7 1 1.4
MS-DRG CC/MCC (%)
Outpatient (0.201) 0.2 (0.701) 0.7 (1) 1 (1.4) 1.4
With (59) 58.8 (52.5) 52.4 (70.5) 70.3 (70.7) 70.6
Without (40.8) 40.7 (46.8) 46.8 (28.5) 28.4 (27.9) 27.8
NA 0.3 0.1 0.4 0.3
ELIX 2.32 (1.84) 2.06 (1.66) 2.33 (1.93) 2.55 (2.03)
CCS 2.06 (2.27) 1.82 (2.05) 2.13 (2.41) 2.05 (2.33)
Region (%)
East North Central 4.7 10.8 12 12.6
East-South-Central 2.8 5.2 5.2 8.7
Middle-Atlantic 23.2 16.8 14.7 13
Mountain 2.8 5.9 4.3 4.6
New England 2.6 3.5 2.8 4
Pacific 4.4 7.4 13.7 10.9
South-Atlantic 52.5 37.7 31.4 29.8
West North Central 1.2 3.4 6.1 5.7
West South Central 6 9.4 9.8 10.7
Hospital size (%)
000–099 3.2 1.5 3.6 4.3
100–199 3.3 7.3 10 13.5
200–299 14 16.9 16 17
300–399 7.5 21.4 21.7 19.2
400–499 9.1 13.4 15.6 14.4
500+ 62.9 39.5 32.9 31.7
Teaching hospital
Yes 66.2 41.2 42.1 40.9
Urban
Yes 95.4 95.4 90.4 89.2

285Journal of Robotic Surgery (2021) 15:275–291
1 3
robotic to non-robotic groups: era-I 2.34, EA 1.35, EM
1.58, LM 1.49, L 1.36). There was also a clear shift toward
smaller hospitals by bed count. Geographically, era-I was
overwhelmingly in the middle-Atlantic (96.8%), then spread
across all regions over time. By era-L, the most dispropor-
tionately RAS regions were middle-Atlantic (14.9% robotic
vs 8.8% non-robotic), east-south Central (12.8 vs 5.7%), and
New England (5.3 vs 2.7%; Table2).
Table 4 (continued)
APR-DRG all patient refined diagnosis related group, BMI body mass index, CCS Charlson Comorbidity Score, ELIX Elixhauser, MS-DRG CC/
MCC medicare severity diagnosis related groups complication or comorbidity (CC) or a major complication or comorbidity (MCC) , SD stand-
ard deviation
Fig. 1 Robotic vs non-robotic
discharge volumes throughout
the study period. a Prostate.
b Uterus. c Colorectal. Solid
black adjusts for hospitals and
ICD9, dashed adjusts for ICD9
only, dotted is unadjusted.
Asterisk, addition of medicare
codes for robotic-assisted
surgery; double asterisk, United
States Preventive Services Task
Force (USPSTF) recommenda-
tion against prostate-specific
antigen (PSA) screening in
prostate cancer patients; double
asterisk, addition of ICD-9 17.4
codes for robotic-assisted sur-
gery; and triple asterisk USP-
STF recommendation against
PSA screening in all patients

286 Journal of Robotic Surgery (2021) 15:275–291
1 3
Fig. 2 Diffusion of robotic eras.
a Prostate, b uterus, c colorectal

287Journal of Robotic Surgery (2021) 15:275–291
1 3
Adoption ofRAS inuterus procedures
Uterus surgeries spanned all five adoption eras as follows:
era-I from 2001Q3 to 2006Q2 (60months, n = 308 dis-
charges, 0.0781% RAS utilization), EA from 2006Q3 to
2008Q1 (21months, n = 1756, 1.13%), EM from 2008Q2 to
2010Q1 (24months, n = 12,546, 6.81%), LM from 2010Q2
to 2012Q2 (27months, n = 49,066, 20.5%), and L from
2012Q3 to 2015Q2 (36months, n = 94,679, 29.0%). Rela-
tive to prostate, uterus spent much longer in era-I (60 vs
15months) and did not enter the EA era until prostate was
well into EM. However, uterus spent a much shorter time in
Majority era relative to prostate (51months from 2008 to
2012 vs 75months from 2005 to 2011. Percent utilization
during the era-L exhibited a decline from a peak of 23.0% in
2013Q4 to 19.1% in 2015Q2 (Table3; Fig.2b).
The mean age of patients in the RAS group was higher
across all eras with the difference being greatest in era-EA
(mean age differences: era-I 0.9, EA 2.8, EM 1.6, LM 1.1, L
1.0years). RAS in era-I was 3.89 times more likely to occur
in Whites compared to Blacks and Hispanics (OR 3.89)
compared to later eras (EA 1.20, EM 1.43, LM 1.31, L 1.38).
Commercial and Medicare comprised between 75 and
90% of discharges across all eras. RAS era-I was biased to
commercial (OR 1.08) compared to non-robotic procedures,
while the era-EA transition shifted toward Medicare (OR
0.737). Subsequent eras gradually saw an attenuated rebal-
ancing toward commercial payers (EM 0.848, LM 0.978, L
0.952).
The presence of RAS inpatient complications was much
lower in era-I (OR 0.551) compared to later eras (EA 1.11,
EM 0.964, LM 0.991, L 1.05). All Patient Refined–Diagno-
sis-Related Group (APR-DRG) mortality scores were much
lower in era-I (OR 1.34 of mild vs moderate/major/extreme)
compared to era-EA (OR 0.603), and lesser to later eras (EM
0.937, LM 0.975, L 0.852).
For hospital demographics, there were significantly fewer
teaching hospitals in era-I compared to later eras, and par-
ticularly era-EA (RR of teaching in robotic vs non-robotic
group: I 0.723, EA 1.55, EM 1.32, LM 1.17, L 0.984).
Uterus era-EA coincided with prostate eras EM and LM
(2006–2008) where RR of teaching hospitals was similarly
elevated. Correspondingly, there was a shift toward larger
hospitals during era-EA by bed count. Geographically, era-I
was predominantly Pacific (64.3%), then spread across all
regions over time (Fig.2b; Table3).
Adoption ofRAS incolorectal procedures
Colorectal spanned just two eras: era-I from 2001Q4 to
2011Q4 (135months, n = 2428, 0.548% RAS utilization),
and EA from 2012Q1 to 2015Q2 (42months, n = 9072,
4.53%). Relative to prostate and uterus, colorectal RAS
procedures remained in era-I for a far longer time (135 vs 15
vs 60months) before experiencing enough growth to enter
EA. Overall low percent utilization suggests that colorectal
procedures may still be in the EA era. As of 2015Q2, colo-
rectal had 7.5% robotic utilization (Table4; Fig.2c).
The mean age of patients in the RAS group was less
in both eras (mean age differences: era-I −1.8, and EA
−1.4years). RAS in era-I was 1.37 times more likely to
occur in Whites compared to Blacks and Hispanics (OR
1.37) than in era-EA (OR 1.20).
Commercial and Medicare comprised between 85 and
90% of discharges across both eras with an overall RAS
bias toward commercial. Whereas both uterus and prostate
procedures experienced an increase in Medicare during the
transition from era-I to -EA, colorectal experienced the
opposite where commercial preference grew from era-I to
era-EA (ORs: era-I 1.22, EA 1.56).
The presence of RAS inpatient complications decreased
from era-I (OR 0.585) to era-EA (0.443). Similarly, APR-
DRG mortality scores were lower in era-I (OR 1.61 of mild
vs moderate/major/extreme) and increased in era-EA (OR
2.42). ELIX and CCS scores were also increased from era-I
to era-EA (mean score differences robotic vs non-robotic:
ELIX era-I −0.09, EA −0.37; CCS era-I − 0.14, EA
−0.28).
For hospital demographics, there were more teaching hos-
pitals in era-I compared to era-EA (RR of teaching in robotic
to non-robotic: era-I 1.57, EA 1.01). Correspondingly, there
was a shift toward smaller hospitals from era-I to era-EA by
bed count. Geographically, era-I was biased toward South-
Atlantic (52.5% robotic vs 31.4% non-robotic) and mid-
Atlantic (23.2 vs 14.7%). era-EA saw a general spread of
RAS, with disproportionate growth in Mountain (Table4).
Discussion
This study, for the first time, provides evidence regarding the
utilization trends, adoption patterns, and factors affecting the
use of RAS both within and across three different surgical
anatomies—prostate, uterus, and colorectal—between 2001
and 2015 throughout the US using Premier database. The
current study observed marked shifts in patient profiles over
time adopting RAS. Prostate and uterus, two surgical anato-
mies with high robotic utilization, demonstrated adoption
rate growth patterns that adhere well to the Rogers Diffu-
sion of Innovations Theory describing the velocity at which
a novel technology diffuses into society, as well as distinct
adopter categories responsible for adoption growth.
A general downtrend in the overall volume of prostate
surgeries observed in the study can be possibly attributed to
a general drop in prostatectomies following a United States
Preventive Services Task Force (USPSTF) recommendation

288 Journal of Robotic Surgery (2021) 15:275–291
1 3
against prostate-specific antigen (PSA) screening for men
at risk for prostate cancer in 2008, and for all men in 2012
[29]. Prostate robotic procedures grew significantly in the
mid-2000s and accounted for over 50% of all prostate dis-
charges starting around 2009. Similar contours of volume
were described by the approach in another study [30].
Similar to the prostate surgeries, a general downtrend in
the overall volume of uterus surgeries was observed, which
tracks the general drop in hysterectomies from its peak
around 2002. A similar trend was observed in a study by
Wright etal. in which overall, 36.4% fewer hysterectomies
were performed in 2010 compared to 68.9% in 2002 [31].
In contrast, prostate surgeries where RAS market saturation
settled at ~ 77% of all hospital discharges, uterus surgeries
stabilized at ~ 29% of all discharges. This was due to both a
greater diversity of uterine procedures and selectivity of sur-
geries utilizing RAS, particularly hysterectomies and uter-
ine resections. The relative utilization of robotic colorectal
procedures was still low. The factor limiting the adoption of
robotics in colorectal procedures could be its cost. Halebi
etal. showed a significantly higher hospital cost associated
with the use of robotic-assisted colonic and rectal resections
[32].
RAS in the prostate procedures spent less time in the era-I
(15months) compared to uterus procedures (60months).
However, the time between the start of EM to the end of
LM was shorter for uterus (4years) compared to the pros-
tate (6years) procedure. This could be due to the fact that
prostate surgeries were the first adopting robotics [33] and
thus, its velocity of adoption was relatively slower than
in the uterus where adoption was driven by surgeons who
were already familiar with the robotic platform through prior
prostate surgeries.
Contrary to prostate and uterus procedures, robotic-
assisted colorectal procedures spanned only two eras (I and
EA). As of 2015Q2, the lowest utilization was observed for
the colorectal procedures, which suggests that colorectal
procedures may still be in the EA era. The first robotic colo-
rectal procedures [34] were performed as early as uterus and
prostate procedures [33, 35, 36], but did not experience a
similar increased velocity of adoption. Longer (135months)
stay of robotic colorectal procedures in era-I suggests that a
very few surgeons are opting for this novel robotic technique
compared to prostate and uterus procedures in its early days.
It is unclear why colorectal procedures had slower adoption
than prostate and uterus procedures.
In the present study, patient characteristics such as race,
insurance types, comorbidity score, and hospital character-
istics were associated with the use of RAS. All the three
RAS procedures in era-I occurred more in Whites compared
to Blacks and Hispanics than in later eras. A similar trend
was observed wherein black and Hispanic patients were less
likely to undergo robotic-assisted radical prostatectomy and
nephrectomy compared to white patients [37, 38]. Previous
studies have shown that Whites had greater accessibility to
novel medical technology [13, 18–20] and are considered
more affluent and educated healthcare consumers, and tend
to have better health outcomes [39, 40].
Furthermore, commercial and Medicare payers (as
opposed to Medicaid) combined to account for 75–90% of all
discharges across all the eras for the three procedures. The
commercial was the preferred payer for RAS across all eras
for all the procedures; however, the relative share dropped in
the middle eras [41]. Both robotic prostate and uterus proce-
dures shifted toward older and more comorbid populations
when entering era-EA. This was consistent with the inclu-
sion of robotic-assisted radiosurgery (codes G0338-G0340),
provider updated by the Department of Health and Human
Services [42]. This sudden population shift could be in
part due to RAS miscoding and undercoding prior to the
inclusion of these codes. Concordantly, in colorectal RAS,
increased preference in commercial observed from era-I to
era-EA may be due to the transition occurring well after the
adoption of Medicare codes.
Regarding hospital characteristics, in prostate and colo-
rectal RAS, a significant decline in teaching hospitals was
observed after exiting era-I. This could be due to the initial
goal of teaching hospitals to be on the frontier of advanced
technologies. Contrastingly in uterus RAS, there were sig-
nificantly fewer teaching hospitals in era-I. The current study
observations are consistent with previous findings in which
the use of RAS in various procedures changes with urban
location, large hospital size, teaching hospitals, and the spe-
cific US regions [43].
Lack of minimal standards for clinical evidence of surgi-
cal devices has led regulatory bodies to endorse an approach
of total product life cycle evaluation [44]. In addition, to
safely develop and adopt new medical devices, the type of
evidence supporting them must be graded depending on the
device’s stage of development, which can be provided by the
IDEAL-D framework [25, 26].
In the current study, both prostate and uterus RAS were
currently spanning era-L with respective utilization of ~ 77
and ~ 29%. Therefore, it can be asserted that these two pro-
cedures are in Stage 4 (long-term study) of the IDEAL-D.
For successor devices in prostate and uterus procedures,
registries should ensure adequate monitoring of safety
and efficacy to identify underperforming devices early. In
addition, for first-of-its-kind devices, clinical trials can be
conducted within registries. On the other hand, colorectal
surgeries are still in era-EA, signifying Stage 3 (assessment
via randomized control trials or alternatives) of the IDEAL-
D framework. A possible replacement for the currently used
devices can be considered in this particular procedure.
The strength of the study lies in the use of a national
cohort from the Premier Hospital Database and

289Journal of Robotic Surgery (2021) 15:275–291
1 3
robotic-assisted laparoscopic ICD-9 CM code, which is a
validated quality-assured identifier for the RAS. This ret-
rospective population-based analysis has a few limitations.
The main limitation of this study lies in its retrospective
nature and its inherent biases. This includes some limita-
tions pertaining to the Premier database, such as the limited
availability of clinical data and possible bias from coding
inaccuracies. Retrospective reviews are prone to selection
bias, and coding errors exist because of the use of discharge
data [45].
Although the dataset included in this study has the advan-
tage of capturing 20% of all inpatient admissions, these data
were obtained through hospital sampling and may be subject
to sampling bias. However, survey weighting was employed
to mitigate this effect. Finally, the Law of Diffusion of Inno-
vation is traditionally applied to binary choices where a sin-
gle agent (person) has either adopted or refused to adopt a
novel technology. In contrast, the adoption of robotic tech-
nology presents the greater complexity of multiple agents
and choices: the hospitals’ decision to purchase a number of
robots to meet potential demand, the surgeons performing a
percentage of their caseloads utilizing RAS, and the patient
electing to undergo surgery with or without the robot.
Conclusions
This study quantified the process of diffusion, utilization,
and factors affecting RAS over time in the prostate, uterus,
and colorectal procedures. RAS uptake in prostate and
uterus procedures acted according to the Theory of Diffu-
sion of Innovations from 2001 to 2015; colorectal proce-
dures appear to remain in early adoption eras. These find-
ings would be complemented by future studies correlating
adoption rates and patient profiles with clinical outcomes
or the cost-effectiveness of novel medical technologies to
assess how adoption is influenced by the balance of patient-,
clinician-, and provider-centered decision points. The next
step in this body of work is to more rigorously analyze spe-
cific patient, surgeon, and hospital subgroups that are tar-
geted over the adoption lifecycle. Coupled with predictive
modeling, this may optimize target patients, procedures, and
hospitals for the uptake of robotic technology in the future.
Acknowledgements The authors acknowledge Amit Koushik, MS and
Ramu Periyasamy, PhD of Indegene Pvt Ltd. for assistance with the
literature review and medical writing.
Author contributions All authors have contributed equally in the devel-
opment of the manuscript.
Funding Study design, collection, analysis, and interpretation of data,
and medical writing support were funded by Johnson & Johnson.
Compliance with ethical standards
Disclosure Gary Chung is an employee of Johnson and Johnson, Inc
which sponsored this study. Piet Hinoul is an employee of Ethicon, Inc
which sponsored this study. Paul Coplan is an employee of Johnson and
Johnson, Inc which sponsored this study. Andrew Yoo is an employee
of Johnson and Johnson, Inc which sponsored this study.
Conflict of interest The authors declare that they have no conflict of
interest.
Ethics approval All procedures followed were in accordance with the
ethical standards of the responsible committee on human experimen-
tation (institutional and national) and with the Helsinki Declaration
of 1975, as revised in 2000. Informed consent was obtained from all
patients for being included in the study.
Informed consent Premier Healthcare Database is aggregated, dei-
dentified, and compliant with the Health Insurance Portability and
Accountability Act and does not require institutional review board
approval. Formal consent is not required.
References
1. Rogers EM (1983) Diffusion of innovations, 3rd edn. Free Press;
Collier Macmillan, New York, London
2. Leggott KT, Martin M, Sklar D, Helitzer D, Rosett R, Crandall
C, Vagh F, Mercer D (2016) Transformation of anesthesia for
ambulatory orthopedic surgery: a mixed-methods study of a dif-
fusion of innovation in healthcare. Healthcare 4:181–187. https ://
doi.org/10.1016/j.hjdsi .2015.09.003
3. Rizan C, Phee J, Boardman C, Khera G (2017) General surgeon’s
antibiotic stewardship: climbing the Rogers Diffusion of Innova-
tion Curve-Prospective Cohort Study. Int J Surg 40:78–82. https
://doi.org/10.1016/j.ijsu.2017.02.040
4. Sagar SP, Law PW, Shaul RZ, Héon E, Langer JC, Wright JG
(2015) Hey, i just did a new operation!: introducing innovative
procedures and devices within an academic health center. Ann
Surg 261:30–31. https ://doi.org/10.1097/SLA.00000 00000 00074 8
5. Marcus HJ, Payne CJ, Hughes-Hallett A, Marcus AP, Yang G-Z,
Darzi A, Nandi D (2016) Regulatory approval of new medical
devices: cross sectional study. BMJ. https ://doi.org/10.1136/bmj.
i2587
6. Marcus HJ, Payne CJ, Hughes-Hallett A, Gras G, Leibrandt K,
Nandi D, Yang G-Z (2016) Making the leap: the translation of
innovative surgical devices from the laboratory to the operat-
ing room. Ann Surg 263:1077–1078. https ://doi.org/10.1097/
SLA.00000 00000 00153 2
7. Volk H-D, Stevens MM, Mooney DJ, Grainger DW, Duda GN
(2015) Key elements for nourishing the translational research
environment. Sci Transl Med. https ://doi.org/10.1126/scitr anslm
ed.aaa20 49
8. Yaxley JW, Coughlin GD, Chambers SK, Occhipinti S, Samara-
tunga H, Zajdlewicz L, Dunglison N, Carter R, Williams S, Payton
DJ, Perry-Keene J, Lavin MF, Gardiner RA (2016) Robot-assisted
laparoscopic prostatectomy versus open radical retropubic pros-
tatectomy: early outcomes from a randomised controlled phase
3 study. Lancet 388:1057–1066. https ://doi.org/10.1016/S0140
-6736(16)30592 -X
9. Akand M, Celik O, Avci E, Duman I, Erdogru T (2015) Open,
laparoscopic and robot-assisted laparoscopic radical prostatec-
tomy: comparative analysis of operative and pathologic outcomes

290 Journal of Robotic Surgery (2021) 15:275–291
1 3
for three techniques with a single surgeon’s experience. Eur Rev
Med Pharmacol Sci 19:525–531
10. Finkelstein J, Eckersberger E, Sadri H, Taneja SS, Lepor H, Dja-
van B (2010) Open versus laparoscopic versus robot-assisted lapa-
roscopic prostatectomy: the European and US experience. Rev
Urol 12:35–43
11. Elhage O, Challacombe B, Shortland A, Dasgupta P (2015) An
assessment of the physical impact of complex surgical tasks on
surgeon errors and discomfort: a comparison between robot-
assisted, laparoscopic and open approaches. BJU Int 115:274–281
12. Alkhateeb S, Lawrentschuk N (2011) Consumerism and its impact
on robotic-assisted radical prostatectomy: consumerism and its
impact on rarp. BJU Int 108:1874–1878. https ://doi.org/10.1111/
j.1464-410X.2011.10117 .x
13. Juo Y-Y, Mantha A, Abiri A, Lin A, Dutson E (2018) Diffusion
of robotic-assisted laparoscopic technology across specialties: a
national study from 2008 to 2013. Surg Endosc 32:1405–1413.
https ://doi.org/10.1007/s0046 4-017-5822-4
14. Campbell B (2012) How to judge the value of innovation. BMJ
344:e1457–e1457. https ://doi.org/10.1136/bmj.e1457
15. Trajtenberg M (1990) A penny for your quotes: patent citations
and the value of innovations. RAND J Econ 21:172. https ://doi.
org/10.2307/25555 02
16. Hofer MD, Meeks JJ, Cashy J, Kundu S, Zhao LC (2013) Impact
of increasing prevalence of minimally invasive prostatectomy on
open prostatectomy observed in the national inpatient sample
and national surgical quality improvement program. J Endourol
27:102–107. https ://doi.org/10.1089/end.2012.0315
17. Damle A, Damle RN, Flahive JM, Schlussel AT, Davids JS, Stur-
rock PR, Maykel JA, Alavi K (2017) Diffusion of technology:
trends in robotic-assisted colorectal surgery. Am J Surg 214:820–
824. https ://doi.org/10.1016/j.amjsu rg.2017.03.020
18. Gabriel E, Thirunavukarasu P, Al-Sukhni E, Attwood K, Nurkin
SJ (2016) National disparities in minimally invasive surgery for
rectal cancer. Surg Endosc 30:1060–1067. https ://doi.org/10.1007/
s0046 4-015-4296-5
19. Kim SP, Boorjian SA, Shah ND, Weight CJ, Tilburt JC, Han LC,
Thompson RH, Trinh Q-D, Sun M, Moriarty JP, Karnes RJ (2013)
Disparities in access to hospitals with robotic surgery for patients
with prostate cancer undergoing radical prostatectomy. J Urol
189:514–520. https ://doi.org/10.1016/j.juro.2012.09.033
20. Monn MF, Bahler CD, Schneider EB, Sundaram CP (2013)
Emerging trends in robotic pyeloplasty for the management of ure-
teropelvic junction obstruction in adults. J Urol 189:1352–1357.
https ://doi.org/10.1016/j.juro.2012.10.001
21. Bouquet de Joliniere J, Librino A, Dubuisson J-B, Khomsi F, Ben
Ali N, Fadhlaoui A, Ayoubi JM, Feki A (2016) Robotic surgery in
gynecology. Front Surg. https ://doi.org/10.3389/fsurg .2016.00026
22. Barkun JS, Aronson JK, Feldman LS, Maddern GJ, Strasberg
SM (2009) Evaluation and stages of surgical innovations. Lancet
374:1089–1096. https ://doi.org/10.1016/S0140 -6736(09)61083 -7
23. Sodergren MH, Darzi A (2013) Surgical innovation and the intro-
duction of new technologies. Br J Surg 100:12–13. https ://doi.
org/10.1002/bjs.9164_5
24. Centers for Medicare & Medicaid Services (CMS) (2008) Medi-
care contractor annual update of the international classification of
diseases, ninth revision, clinical modification (ICD-9-CM). https
://www.cms.gov/Outre ach-and-Educa tion/Medic are-Learn ing-
Netwo rk-MLN/MLNMa tters Artic les/downl oads/MM610 7.pdf
25. Sedrakyan A, Campbell B, Merino JG, Kuntz R, Hirst A, McCull-
och P (2016) IDEAL-D: a rational framework for evaluating
and regulating the use of medical devices. BMJ. https ://doi.
org/10.1136/bmj.i2372
26. Hirst A, Philippou Y, Blazeby J, Campbell B, Campbell M,
Feinberg J, Rovers M, Blencowe N, Pennell C, Quinn T, Rog-
ers W, Cook J, Kolias AG, Agha R, Dahm P, Sedrakyan A,
McCulloch P (2018) No surgical innovation without evalua-
tion: evolution and further development of the IDEAL frame-
work and recommendations. Ann Surg. https ://doi.org/10.1097/
SLA.00000 00000 00279 4
27. Dahm P, Sedrakyan A, McCulloch P (2014) Application of
the IDEAL framework to robotic urologic surgery. Eur Urol
65:849–851. https ://doi.org/10.1016/j.eurur o.2013.11.003
28. Menon M, Sood A, Bhandari M, Kher V, Ghosh P, Abaza R,
Jeong W, Ghani KR, Kumar RK, Modi P, Ahlawat R (2014)
Robotic kidney transplantation with regional hypothermia:
a step-by-step description of the Vattikuti Urology Institute-
Medanta technique (IDEAL Phase 2a). Eur Urol 65:991–1000.
https ://doi.org/10.1016/j.eurur o.2013.12.006
29. Halpern JA, Shoag JE, Artis AS, Ballman KV, Sedrakyan A,
Hershman DL, Wright JD, Shih YCT, Hu JC (2017) National
trends in prostate biopsy and radical prostatectomy volumes fol-
lowing the US preventive services task force guidelines against
prostate-specific antigen screening. JAMA Surg 152:192. https
://doi.org/10.1001/jamas urg.2016.3987
30. Leow JJ, Chang SL, Trinh Q-D (2016) Accurately determin-
ing patients who underwent robot-assisted surgery: limitations
of administrative databases. BJU Int 118:346–348. https ://doi.
org/10.1111/bju.13423
31. Wright JD, Herzog TJ, Tsui J, Ananth CV, Lewin SN, Lu Y-S,
Neugut AI, Hershman DL (2013) Nationwide trends in the per-
formance of inpatient hysterectomy in the United States. Obstet
Gynecol 122:233–241. https ://doi.org/10.1097/AOG.0b013
e3182 99a6c f
32. Halabi WJ, Kang CY, Jafari MD, Nguyen VQ, Carmichael JC,
Mills S, Stamos MJ, Pigazzi A (2013) Robotic-assisted colo-
rectal surgery in the united states: a nationwide analysis of
trends and outcomes. World J Surg 37:2782–2790. https ://doi.
org/10.1007/s0026 8-013-2024-7
33. Davies BL, Hibberd RD, Ng WS, Timoney AG, Wickham JEA
(1991) The development of a surgeon robot for prostatectomies.
Proc Inst Mech Eng [H] 205:35–38. https ://doi.org/10.1243/
PIME_PROC_1991_205_259_02
34. Weber PA, Merola S, Wasielewski A, Ballantyne GH (2002)
Telerobotic-assisted laparoscopic right and sigmoid colectomies
for benign disease. Dis Colon Rectum 45:1689–1694. https ://
doi.org/10.1097/01.DCR.00000 37657 .78153 .A8 (discussion
1695–1696)
35. Falcone T, Goldberg JM, Margossian H, Stevens L (2000)
Robotic-assisted laparoscopic microsurgical tubal anastomosis:
a human pilot study. Fertil Steril 73:1040–1042
36. Binder J, Jones J, Bentas W, Wolfram M, Bräutigam R, Probst
M, Kramer W, Jonas D (2002) Robot-assisted laparoscopy in
urology. Radical prostatectomy and reconstructive retroperito-
neal interventions. Urol Ausg A 41:144–149
37. Kim J, ElRayes W, Wilson F, Su D, Oleynikov D, Morien M,
Chen L-W (2015) Disparities in the receipt of robot-assisted
radical prostatectomy: between-hospital and within-hospital
analysis using 2009-2011 California inpatient data. BMJ Open
5:e007409–e007409. https ://doi.org/10.1136/bmjop en-2014-
00740 9
38. Alameddine M, Koru-Sengul T, Moore KJ, Miao F, Sávio
LF, Nahar B, Prakash NS, Venkatramani V, Jue JS, Punnen S,
Parekh DJ, Ritch CR, Gonzalgo ML (2018) Trends in utiliza-
tion of robotic and open partial nephrectomy for management
of cT1 renal masses. Eur Urol Focus. https ://doi.org/10.1016/j.
euf.2017.12.006
39. Tseng TY, Cancel QV, Fesperman SF, Kuebler HR, Sun L, Rob-
ertson CN, Polascik TJ, Moul JW, Vieweg J, Albala DM, Dahm
P (2007) The role of early adopter bias for new technologies in
robot assisted laparoscopic prostatectomy. J Urol 177:1318–1323.
https ://doi.org/10.1016/j.juro.2006.11.035

291Journal of Robotic Surgery (2021) 15:275–291
1 3
40. Scandlen G (2004) Commentary-how consumer-driven health
care evolves in a dynamic market: commentary-how consumer-
driven health care evolves in a dynamic market. Health Serv Res
39:1113–1118. https ://doi.org/10.1111/j.1475-6773.2004.00276
.x
41. Lienig J, Bruemmer H (2017) Fundamentals of electronic systems
design. Springer, Berlin
42. Department of Health and Human Services (2003) Medicare
program; revisions to payment policies under the physician fee
schedule for calendar year 2004; final rule
43. Cooper MA, Hutfless S, Segev DL, Ibrahim A, Lyu H, Makary
MA (2014) Hospital level under-utilization of minimally inva-
sive surgery in the United States: retrospective review. BMJ
349:g4198–g4198. https ://doi.org/10.1136/bmj.g4198
44. Pennell CP, Hirst A, Sedrakyan A, McCulloch PG (2016) Adapt-
ing the IDEAL framework and recommendations for medical
device evaluation: a modified Delphi survey. Int J Surg 28:141–
148. https ://doi.org/10.1016/j.ijsu.2016.01.082
45. Lorence DP, Ibrahim IA (2003) Benchmarking variation in cod-
ing accuracy across the United States. J Health Care Finance
29:29–42
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.