Keywords
conflict of interest - health policy - industry payments
While financial relationships between physicians and manufacturers of pharmaceuticals
and medical devices have long existed, these relationships have recently come under
increasing scrutiny due to concerns of impact on physicians' prescribing patterns.
To address these concerns, the Physician Payments Sunshine Act was enacted in 2010,
requiring manufacturers to report to the Centers for Medicare and Medicaid Services
(CMS) any payments made to physicians totaling over $100 per year.[1] It also required the creation of the Open Payments Program, which makes these data
publicly available.[2]
The improved transparency from this program has facilitated initial investigations
into who receives money from industry, as well as implications of these payments.
Several studies have examined the relationship between industry payments and research
output, revealing a connection between receiving payments and increased scholarly
impact.[3]
[4]
[5]
[6]
[7] In the field of ophthalmology, industry payments have been associated with distinct
patterns in prescribing of antivascular endothelial growth factor (VEGF) agents, prompting
questions about the impact of these payments on ophthalmologists.[8]
[9]
To broaden the understanding of industry payments to ophthalmologists, we conducted
a comprehensive analysis of Open Payments data for every academic ophthalmologist
in the United States, examining associations between industry support and scholarly
impact. We also analyzed the influence of subspecialty, gender, academic rank, type
of degree, and additional degree(s) (e.g., MBA, MPH) on scholarly output. As ophthalmologists
with high scholarly impact are considered academic leaders and are often asked to
serve on guideline committees or lead practice-determining clinical trials, an understanding
of these dynamics is a critical step toward increasing transparency and research integrity
in ophthalmology.
Methods
Study Sample
Our sample consisted of ophthalmology faculty included on publicly available web sites
of all 117 accredited ophthalmology residency programs in the continental United States,
as listed by the Accreditation Council for Graduate Medical Education.[10] Using the web sites for each of these residency programs, the following information
was extracted for each ophthalmologist: name, gender, primary medical degree (MD,
DO, MD/PhD), presence of additional degree(s) (binary variable; included degrees such
as MBA, MPH, and MHS), and academic rank (assistant professor, associate professor,
full professor). Subspecialty data were collected from the American Academy of Ophthalmologists
(AAO) “Find an Ophthalmologist” database and from faculty webpages.[11]
We excluded faculty without an MD or DO degree, as reporting requirements apply to
physicians only.[12] We also excluded ophthalmologists whose title was not one of the following: assistant
professor, associate professor, or professor. Ophthalmologists whose appointment specified
a primarily clinical role (e.g., “Clinical Associate Professor”) were excluded, as
we assumed that research may not be a primary component of their responsibilities.
We also excluded ophthalmology faculty members from military training programs and
from programs located outside of the continental United States. Ophthalmologists whose
most recent publication was more than 10 years before 2016 were also excluded, because
we assumed that these ophthalmologists were no longer actively involved in research.
Finally, ophthalmologists for whom subspecialty data could not be located were excluded
from the subspecialty analysis only.
Payments to Physicians
Industry payments to ophthalmologists were identified using the CMS Open Payments
Database for the year 2016,[2] which was the most recent year available at the time of data collection for this
study. The Open Payments Database makes information about financial relationships
between physicians and certain healthcare manufacturers publicly available. Payments
of the following types were recorded for each ophthalmologist: general payments (i.e.,
payments not associated with research), research payments (i.e., payments associated
with a research agreement or protocol), physician ownership or investment interest
(i.e., value of ownership or investment interest in certain healthcare manufacturers),
and associated research funding (i.e., payments to a research institution that name
the physician as a principal investigator).[13] Total payments received by each ophthalmologist were divided into the following
payment categories, consistent with those used in a previous analysis of the Open
Payments data: less than $100, $100 to 1,000, $1,000 to 10,000, $10,000–100,000, and
greater than $100,000.[4] General payments were also grouped into the following types: food and beverage,
education, consulting, speaker at a Continuing Education Program (CEP), speaker at
a non-CEP, travel and lodging, honorarium, entertainment, gifts, and other (e.g.,
charitable contributions, royalty, or license fees).
Outcome Variables
To assess the overall scholarly impact of each ophthalmologist, bibliometric data
were extracted using the “Author Search Tool” from Scopus.[14] Scopus is a database of peer-reviewed literature with information about citations.
The following data were extracted for each ophthalmologist: Hirsch index (H-index),
total number of documents, total citations, and range of publication activity (in
years). H-index is a validated marker of scholarly impact, defined by having published
h publications with at least h citations per publication.[15] For example, a physician with an H-index of 30 has published at least 30 publications
with at least 30 citations each. This metric has been shown to be a better measure
of the overall quality and quantity of a scientist's research than other variables,
such as total citations or total publications.[16]
Statistical Analysis
Descriptive statistics were used to characterize the sample. We used analysis of variance
and two-sided unpaired t-tests for comparison of continuous variables. A p-value of 0.05 was used for all statistical tests. All analyses were performed using
Stata 15.0 and outputs were displayed with GraphPad Prism 7.0. Institutional Review
Board approval was obtained from the Yale University Institutional Review Board.
Results
Academic Ophthalmologists
There were 1,653 academic ophthalmologists that met inclusion criteria for our study
([Table 1]). Of these ophthalmologists, 1,104 (67%) were men and 549 (33%) were women. Academic
rank was assistant professor for 649 (39%), associate professor for 392 (24%), and
full professor for 612 (37%). The most common subspecialties were surgical retina
(17%), glaucoma (14%), cornea/external disease (13%), and pediatric ophthalmology/strabismus
(11%). Additional degrees such as MPH and MBA were identified for 140 (8%) academic
ophthalmologists.
Table 1
Characteristics of academic ophthalmologists receiving industry payments and amount
of payments received
|
Ophthalmologist characteristic
|
Number of ophthalmologists (% of total)
|
Number of ophthalmologists paid (% of category)
|
Mean payment (standard error of mean), in thousands of dollars [a]
|
|
Gender
|
|
Male
|
1,104 (67)
|
842 (76)
|
34.7 (5.7)
|
|
Female
|
549 (33)
|
383 (70)
|
15.1 (5.7)
|
|
Academic rank
|
|
Assistant professor
|
649 (39)
|
478 (74)
|
6.7 (1.5)
|
|
Associate professor
|
392 (24)
|
283 (72)
|
22.2 (3.9)
|
|
Full professor
|
612 (37)
|
464 (76)
|
55.0 (11.0)
|
|
Primary Degree
|
|
MD
|
1,404 (85)
|
1,067 (76)
|
29.5 (4.9)
|
|
MD/PhD
|
230 (14)
|
150 (65)
|
22.0 (4.6)
|
|
DO
|
19 (1)
|
8 (42)
|
29.8 (29.6)
|
|
Additional degree(s) [b]
|
|
Yes
|
140 (8)
|
98 (70)
|
29.5 (4.7)
|
|
No
|
1,513 (92)
|
1,127 (74)
|
18.5 (5.3)
|
|
Subspecialty
|
|
Cataract/anterior segment
|
44 (3)
|
40 (91)
|
27.0 (11.4)
|
|
Comprehensive
|
100 (6)
|
77 (77)
|
4.5 (1.8)
|
|
Cornea/external disease
|
208 (13)
|
175 (84)
|
16.2 (4.2)
|
|
Glaucoma
|
229 (14)
|
187 (82)
|
35.5 (12.0)
|
|
Neuro-ophthalmology
|
123 (7)
|
70 (57)
|
30.9 (20.0)
|
|
Ocular Oncology
|
24 (1)
|
17 (71)
|
6.6 (5.2)
|
|
Oculoplastics
|
133 (8)
|
109 (82)
|
2.2 (1.0)
|
|
Ophthalmic pathology
|
25 (2)
|
11 (44)
|
0.5 (0.3)
|
|
Pediatric/strabismus
|
190 (11)
|
106 (56)
|
4.5 (1.8)
|
|
Refractive surgery
|
19 (1)
|
18 (95)
|
15.3 (9.4)
|
|
Medical retina
|
68 (4)
|
52 (76)
|
26.8 (10.5)
|
|
Surgical retina
|
284 (17)
|
243 (86)
|
57.8 (14.1)
|
|
Uveitis/immunology
|
36 (2)
|
29 (81)
|
33.0 (15.0)
|
|
Other
|
33 (2)
|
12 (36)
|
62.9 (39.0)
|
|
Subspecialty not identified
|
137 (8)
|
79 (58)
|
46.7 (34.7)
|
Notes: Authors' analysis of data from open payments database.
a Excludes individuals who did not receive any industry funding.
b Includes MD, DO, and MD/PhD ophthalmologists.
Payments to Academic Ophthalmologists
Of the 1,653 academic ophthalmologists in our sample, 1,225 (74%) received industry
payments in 2016, totaling $35.0 million. The majority of this funding was in the
form of general payments ($18.9 million, 54%) and associated research payments ($13.1
million, 37%); the remaining payments were physician ownership ($2.8 million, 8%)
and direct research payments ($0.2 million, <1%).
Among academic ophthalmologists who received any form of industry payment, the median
payment amount was $265 (interquartile range, $81–4,341), and the mean was $28,582.
Of these ophthalmologists, 352 (29%) received less than $100, 468 (38%) received between
$100 and 1,000, 172 (14%) received between $1,000 and 10,000, 152 (12%) received between
$10,000 and 100,000, and 81 (7%) received greater than $100,000 ([Table 2]). Payments totaling more than $10,000 were received by males more commonly than
females (22 vs13%), and by full and associate professors more commonly than assistant
professors (27 vs 23% vs 8%, respectively). Ophthalmologists in the following subspecialties
most commonly received payments totaling more than $10,000: surgical retina (33%),
glaucoma (26%), uveitis/immunology (24%), cataract/anterior segment (22%), and other
(42%).
Table 2
Distribution of industry payments received by academic ophthalmologists
|
Ophthalmologist characteristic
|
Ophthalmologists receiving payments in each category: No. (%)
|
|
<$100
|
$100–1,000
|
$1,000–10,000
|
$10,000–100,000
|
>$100,000
|
|
All ophthalmologists
|
352 (29)
|
468 (38)
|
172 (14)
|
152 (12)
|
81 (7)
|
|
Gender
|
|
Male
|
222 (26)
|
313 (37)
|
123 (15)
|
115 (14)
|
69 (8)
|
|
Female
|
130 (34)
|
155 (40)
|
49 (13)
|
37 (10)
|
12 (3)
|
|
Academic rank
|
|
Assistant professor
|
171 (36)
|
214 (45)
|
53 (11)
|
31 (6)
|
9 (2)
|
|
Associate professor
|
72 (25)
|
108 (38)
|
37 (13)
|
46 (16)
|
20 (7)
|
|
Full professor
|
109 (23)
|
146 (31)
|
82 (18)
|
75 (16)
|
52 (11)
|
|
Primary degree
|
|
MD
|
317 (30)
|
415 (39)
|
141 (13)
|
123 (12)
|
71 (7)
|
|
DO
|
3 (38)
|
4 (50)
|
0 (0)
|
0 (0)
|
1 (12)
|
|
MD/PhD
|
32 (21)
|
49 (33)
|
31 (21)
|
29 (19)
|
9 (6)
|
|
Additional degree(s)[a]
|
|
Yes
|
23 (23)
|
33 (34)
|
22 (22)
|
15 (15)
|
5 (5)
|
|
No
|
329 (29)
|
435 (39)
|
150 (13)
|
137 (12)
|
76 (7)
|
|
Subspecialty
|
|
Cataract/anterior segment
|
11 (28)
|
14 (35)
|
6 (15)
|
5 (12)
|
4 (10)
|
|
Comprehensive
|
28 (36)
|
40 (52)
|
2 (3)
|
7 (9)
|
0 (0)
|
|
Cornea/external disease
|
55 (31)
|
72 (41)
|
21 (12)
|
16 (9)
|
11 (6)
|
|
Glaucoma
|
36 (19)
|
67 (36)
|
35 (19)
|
35 (19)
|
14 (7)
|
|
Neuro-ophthalmology
|
28 (40)
|
28 (40)
|
6 (9)
|
4 (6)
|
4 (6)
|
|
Ocular oncology
|
6 (35)
|
8 (47)
|
1 (6)
|
2 (12)
|
0 (0)
|
|
Oculoplastics
|
47 (43)
|
52 (48)
|
7 (6)
|
3 (3)
|
0 (0)
|
|
Ophthalmic pathology
|
3 (27)
|
6 (55)
|
2 (18)
|
0 (0)
|
0 (0)
|
|
Pediatric/strabismus
|
42 (40)
|
40 (38)
|
15 (14)
|
8 (8)
|
1 (1)
|
|
Refractive surgery
|
2 (11)
|
7 (39)
|
6 (33)
|
2 (11)
|
1 (6)
|
|
Medical retina
|
20 (38)
|
11 (21)
|
11 (21)
|
5 (10)
|
5 (10)
|
|
Surgical retina
|
39 (16)
|
77 (32)
|
47 (19)
|
47 (19)
|
33 (14)
|
|
Uveitis/immunology
|
7 (24)
|
13 (45)
|
2 (7)
|
4 (14)
|
3 (10)
|
|
Other
|
1 (8)
|
4 (33)
|
2 (17)
|
3 (25)
|
2 (17)
|
Notes: Table showing proportion of academic ophthalmologists receiving different total
payment amounts, and variations in this distribution by ophthalmologist characteristic.
a Includes MD, DO, and MD/PhD ophthalmologists.
Association between Industry Payments and Scholarly Impact
There was no statistically significant difference between the mean H-index of academic
ophthalmologists receiving any industry payments and those not receiving any payments
(p = 0.68). However, among academic ophthalmologists who did receive industry payments,
there was a significant difference in H-index by payment category (p < 0.001) ([Fig. 1]). The mean H-index was 12.6 for ophthalmologists receiving less than $100, 12.2
for those receiving between $100 and 1,000, 18.8 for those receiving between $1,000
and 10,000, 21.3 for those receiving between $10,000 and 100,000, and 29.4 for those
receiving greater than $100,000. These differences remained statistically significant
when analyzing other bibliometric indicators, including total number of publications
(p < 0.001) and total number of citations (p < 0.001).
Fig. 1 Association between industry payments and scholarly impact. Notes: Bar graph showing
authors' analysis of bibliometric data stratified by total industry payments received,
with scholarly impact measured by (top) mean H-index, (middle) mean number of publications,
and (bottom) mean number of citations. Error bars indicate standard error of mean.
In analysis examining general payments only, there was a significant difference in
mean H-index by general payment amount (p < 0.001). The most common general payment type was food and beverage, with 1,100
ophthalmologists (94% of ophthalmologists receiving any general payment) receiving
this type of payment ([Table 3]). Additionally, there was a significant difference between the mean H-index of ophthalmologists
receiving greater than $1,000 versus less than $1,000 in the following general payment
types: food and beverage (p < 0.001), travel and lodging (p < 0.001), non-CEP speaker fees (p = 0.02), and honoraria (p = 0.048).
Table 3
Distribution of general industry payments by payment type and association with H-index
|
General payment type
|
Total payments (dollars)
|
Mean payment (dollars)
|
Number of ophthalmologists paid (% of ophthalmologists receiving general payments)
|
Mean H-index if paid <$1,000
|
Mean H-index if paid >$1,000
|
p-Value
|
|
Food and beverage
|
424,588
|
362
|
1,100 (94)
|
14.4
|
21.9
|
< 0.001***
|
|
Travel and lodging
|
955,217
|
814
|
243 (21)
|
16.9
|
24.4
|
<0.001***
|
|
Consulting
|
4,178,866
|
3,560
|
204 (17)
|
24.7
|
25.3
|
0.89
|
|
Education
|
2,918,780
|
2,486
|
152 (13)
|
14.3
|
23.7
|
0.26
|
|
Speaker fees: non-CEP
|
4,351,819
|
3,707
|
141 (12)
|
16.0
|
26.2
|
0.02*
|
|
Honorarium
|
194,095
|
165
|
33 (3)
|
18.0
|
30.9
|
0.048*
|
|
Other
|
5,615,058
|
4,783
|
32 (3)
|
34.6
|
26.0
|
0.32
|
|
Speaker fees: CEP
|
181,051
|
154
|
28 (2)
|
12.3
|
22.4
|
0.36
|
|
Gifts
|
45,136
|
38
|
26 (2)
|
21.6
|
21.0
|
0.95
|
|
Entertainment
|
4,928
|
4
|
15 (1)
|
22.3
|
26.5
|
0.68
|
Abbreviation: CEP, Continuing Education Program.
Notes: Authors' analysis of general payment data, segregated by general payment type,
and differences in mean H-index between physicians receiving less than versus greater
than $1,000 of payments of each type. Statistical significance indicated by * for
p < 0.05, ** for p < 0.01, and *** for p < 0.001.
Subgroup Analysis
[Fig. 2] summarizes the results of the subgroup analysis by gender, academic rank, degree(s),
and subspecialty. Within each academic rank and gender, industry payments greater
than $1,000 were associated with a higher mean H-index ([Fig. 2]). This trend also held true for ophthalmologists whose only professional degree
was MD and ophthalmologists with additional degrees such as MPH and MBA, but we did
not observe the same dynamic for MD/PhDs. Additionally, this trend held true for ophthalmologists
in the following subspecialties: cornea/external disease, glaucoma, neuro-ophthalmology,
oculoplastics, pediatric ophthalmology/strabismus, and surgical retina.
Fig. 2 Mean H-index of ophthalmologists receiving <$1,000 versus >1,000 of industry payments.
Notes: Mean H-index of academic ophthalmologists receiving less than versus more than
$1,000 in total industry payments, stratified by (top) gender, academic rank, degree(s),
and (bottom) subspecialty. Ophthalmologists whose only degree was DO were excluded
from the degree analysis due to limited sample size (n = 8). Ophthalmologists for whom subspecialty data could not be found were excluded
from the subspecialty analysis. Error bars indicate standard error of mean. Statistical
significance indicated by *for p < 0.05, **for p < 0.01, and ***for p < 0.001.
Discussion
In this comprehensive analysis of academic ophthalmologists in the United States,
we found a statistically significant association between industry payments and scholarly
impact among ophthalmologists who received these payments. This finding held true
even after controlling for factors such as gender, academic seniority, and subspecialty.
Additionally, analysis limited to nonresearch payments showed a similarly strong association.
While our findings do not indicate causality, we demonstrate a strong association
between industry funding and scholarly impact among academic ophthalmologists.
Our findings are consistent with research showing similar associations between industry
payments and scholarly impact in other specialties, including otolaryngology,[6] plastic surgery,[4] neurosurgery,[5] transplant surgery,[3] and oncology.[7] While research in ophthalmology has previously shown that H-index varies by academic
rank and fellowship,[17] our study is the first to investigate the impact of industry funding on scholarly
impact. In addition to confirming these differences in scholarly impact by academic
rank and subspecialty, our analysis also extends these findings to show that industry
funding modulates these associations. Within subspecialties such as cornea/eternal
disease, glaucoma, oculoplastics, pediatric ophthalmology/strabismus, and surgical
retina, ophthalmologists receiving greater than $1,000 in industry funding had significantly
higher mean H-index than those receiving less than $1,000. We did not observe these
differences in other subspecialties, likely due to limited sample size. For example,
only 11 of the 25 ophthalmic pathologists in our sample received payments, and only
two of these received greater than $1,000 in industry payments; ophthalmic pathologists
may also be trained through pathology rather than ophthalmology. Our findings are
also consistent with research about gender differences in industry funding of ophthalmologists,
again demonstrating that female ophthalmologists receive industry payments considerably
less commonly than men and in lower quantities.[18]
[19] Finally, our analyses mirror the results of a previous study that identified an
association between National Institutes of Health (NIH) funding and H-index of ophthalmologists.[20]
There are several possible explanations for the association between industry funding
and scholarly impact among academic ophthalmologists. Pharmaceutical companies may
seek influential ophthalmologists (often referred to as “key opinion leaders”) for
speaking and consulting engagements, preferentially making payments to ophthalmologists
with high scholarly impact. Additionally, ophthalmologists' research may benefit indirectly
from industry payments through access to other academic leaders. For example, an ophthalmologist
who is paid for a speaking engagement may meet potential research collaborators and
gain exposure to new research ideas, which could ultimately lead to increased scholarly
impact. Finally, given that consulting fees and travel/lodging represented two of
the largest categories of general payments in our analysis, it is possible that ophthalmologists
involved in consulting engagements benefit from increased access to research opportunities
in collaboration with industry researchers, ultimately increasing their scholarly
output.
Given the growing scrutiny of the pharmaceutical industry's impact on medicine, our
findings have important implications for the field of academic ophthalmology. Many
academic publications and general media outlets have focused on the ways that the
pharmaceutical industry can impact prescribing,[21] disease definitions,[22] healthcare expenditures,[23] and medical education.[24]
[25] In ophthalmology, industry payments have been shown to impact VEGF prescribing;[8]
[9] however, a study of ophthalmology residents showed that while the majority of trainees
interact with pharmaceutical representatives, these trainees commonly feel that they
are not susceptible to the influence of pharmaceutical promotions.[26] This discrepancy suggests that clinicians may be subconsciously influenced by industry
contributions, and warrants further investigation. Further, industry payments may
influence patient perceptions of ophthalmologists: a recent study showed that patients
commonly disapprove of their ophthalmologists receiving $100 or more in industry payments
of any kind.[27] In our study, the majority of ophthalmologists received this level of payment.
Many professional organizations, including industry organizations such as the Pharmaceutical
Research and Manufacturers of America[28] and physician organizations such as the AAO, have responded to these concerns by
publishing guidance on relationships between industry and healthcare professionals
and by promoting transparency through requiring conflict of interest (COI) disclosures.
For example, AAO requires that anyone who is in a “position to influence content”
must disclose their COIs.[29] This policy requires individuals such as speakers and reviewers to disclose if they
have a “financial relationship with a commercial interest and the opportunity to affect
the content of CME (Continuing Medical Education) about the products or services of
that commercial interest.” However, to our knowledge, there are no reporting requirements
for payments that are unrelated to an individual's current AAO engagement but that
may have influenced that individual's scholarly productivity, such as the nonresearch
payments analyzed in our study.
Our findings also have implications for career development in ophthalmology, given
the emphasis on producing high-impact scholarly work to achieve academic career progression.
Because metrics such as number of publications and citation frequency have historically
been used to support academic promotion decisions, knowledge of how industry funding
impacts these outcomes is an essential component of understanding career development
in academic ophthalmology.
There are several limitations to our study. First, our study was not designed to identify
a causal relationship between industry funding and scholarly impact. Our study examines
an association between these factors, and therefore cannot determine whether industry
payments result in higher H-index or whether an ophthalmologist's research productivity
leads to greater industry payments, or a combination of both. Second, while the H-index
is a widely used surrogate for scholarly productivity, it has several limitations:
it does not give credit to heavily cited articles, it is not an optimal tool for making
comparisons across different types of research (e.g., basic science research vs retrospective
epidemiological research), it favors researchers who publish in developed fields and
clinical trials, and it can be artificially inflated by self-citation.[30] Third, the Scopus database occasionally splits a single author's information into
multiple profiles, often after an author switches institutions. In these cases, we
chose the profile with the highest H-index, although this approach still underestimates
the author's scholarly productivity. Finally, our analysis was limited to industry
support and did not control for other sources of funding, such as NIH grants, which
are associated with greater scholarly impact.[20]
In summary, our analysis demonstrates a significant association between industry payments
and scholarly impact, even after adjusting for gender, academic rank, and subspecialty.
While our study cannot identify a causal relationship, our findings do highlight the
important role of industry funding in academic ophthalmology research. As the Open
Payments Program accumulates additional data and COI transparency increases, future
research should further characterize the relationship between industry funding and
scholarly impact in the field of ophthalmology.
