Real World Trends in Intravitreal Injection Practices Among American Retina Specialists

Abstract

Purpose:

To analyze practice patterns employed for intravitreal injections (IVI) by retinal specialists in the United States.

Design:

Cross-sectional online survey.

Participants:

Retina specialists in the United States who responded to a web-based survey.

Methods:

Retinal specialists in the United States were contacted via email to complete a web-based, anonymous 24-question survey. Multivariate analysis was performed on a selected question of interest focused on choice of anesthetic used for IVI.

Main Outcome Measures:

Differences in IVI practices, such as antibiotic preferences, and different odds of anesthetic use by demographic variables with 95% confidence intervals (CI).

Results:

A total of 281 retinal specialists responded to the survey (17% response rate). Respondents had an average age of 53 years with an average of 20 years in practice. Respondents practiced in 42 states, with 90% practicing in an urban or suburban area. For anesthesia, 14% utilized a topical anesthetic with cotton swab compression, 27% a subconjunctival anesthetic, and 31% an anesthetic gel. Age, gender, geographic location, and practice setting did not appear to significantly impact choice of anesthetic for IVI. 66% of respondents always used a lid speculum, 21% administered topical antibiotics prior to injection, 36% wore a mask, 73% wore gloves, and 45% always dilated the eyes prior to injection. The majority of respondents utilized a 30-gauge needle and injected in the inferior temporal quadrant (70%). 45% would always perform bilateral injections the same day if indicated. After the injection, 14% administered post op NSAIDs, 28% administered post-op antibiotics, and 31% routinely checked intraocular pressure after injection.

Conclusions:

This study provides real-world trends in practices for IVI among retina specialists in the United States. In addition, age, gender, practice type, or geographic location did not influence anesthetic choice for IVI.

Precis

This large survey of intravitreal injection practices among retina specialists in the United States illustrates current real world trends and preferences among retina specialists.

Introduction

Intravitreal Injections (IVI) have now been accepted as the standard of care for delivering therapeutics to the retina. Since 2006, there has been rapid increase in the number of intravitreal injections, estimated to have been around 5.9 million in 2016.¹ There is a wide spectrum of approaches employed for preparation, technique, antibiotic usage, and post injection examinations for intravitreal injections.^2,3 These practices generally involve considerations for preventing endophthalmitis, improving patient comfort, and optimizing clinical workflow. The most concerning risk associated with IVI has been the development of endophthalmitis. Employing different techniques before, during and after IVI procedures, such as the decision to use povidone-iodine either in addition to or instead of prophylactic antibiotics, can affect the rates of endophthalmitis.⁴ Additional complications of IVI include an elevated intraocular pressure, ocular hemorrhage, rhegmatogenous retinal detachment, and cataract formation.^4–7

While various guidelines have been published by various groups for evidence based practices surrounding IVI, it is unclear how these practices are incorporated in the real world.^2,3 Furthermore it is unknown whether demographic characteristics, such as age, gender, or practice setting can affect certain practices, such as choice of anesthetic used for IVI. The purpose of this study is to survey current practices surrounding intravitreal injections in the United States and to analyze differences based on various demographic parameters for choice of anesthetic used in IVI.

Methods

A subset of retina specialists in the United States were contacted via email to complete a cross-sectional 24 question, internet-based survey aimed at better understanding pre-procedure methodologies, materials used, medications used, and post-procedure protocols related to IVI. This subset was identified by cross referencing names from the American Society of Retina Specialists Find a Retina Specialist webpage (https://www.asrs.org/find-a-specialist) (USA as a search term)n to contact information obtained from publicly available sources as well as other retina specialists.

The survey was created in the Qualtrics Research Core Platform, a web-based platform that allows data storage in a password-protected database and export data into CSV, PDF, and Powerpoint formats. The survey was entirely anonymous and thus no identifiable information was acquired or stored in the database and is compliant with current HIPAA guidelines. The first page of the survey was an informed consent page and required every participant to agree to the informed consent to be eligible to complete the survey. This research study has been certified as exempt from IRB review by the UC San Diego Human Research Protections Program and permits the administration of the web-based survey to retinal specialists. The research study adhered to the tenets of the Declaration of Helsinki.

On November 5^th, 2018 , emails were sent out to 1676 retina specialists in the United States, of which 34 emails immediately were returned. Thus, the survey was successfully sent out to 1642 participants, of which 281 completed the survey. Between November 5^th and December 11^th, two email reminders to complete the survey were sent to the retina specialists. The second email reminder stated that the survey would be closed within 7 days. The survey response collection was stopped on December 14^th, 2018 and results were analyzed on December 17^th, 2018.

For choice of anesthetic used for IVI, we used logistic regression to quantify and evaluate the relationship between demographic covariates and responses. Response to a question was treated as a binary (yes/no) with non-responders excluded from the analysis. All demographic covariates of interest were included in the multivariate regression, which include age, gender, geographic location, and practice setting. We reported odds ratios with relation to reference groups as discussed in the results section, as well as 95% confidence intervals and p-values using the Wald test. All statistics were performed using R version 3.5.2 (https://cran.r-project.org/).”

Results

Demographics

There were a total of 281 individuals (17% response rate) who responded to the survey by the allotted deadline, which was stated in the final reminder email. Some of the respondents chose not to respond to select questions in the survey. The age range of respondents was 29 to 78, with an average age of 53. Eighty-five percent of respondents identify as male (240/281) and 13% as female (37/281). Among the respondents, 11% have been in practice between 0 and 5 years post-fellowship (31/279), 14% between 6 and 10 (39/279), 11% between 11 and 15 (32/279), 15% between 16 and 20 (41/279), and 49% for 21 or more years (136/279), with an average practice length of 20 years post-fellowship and practice length range between 1 and 46 years. In terms of geographic distribution, physicians from 42 different states responded to the survey, with the largest representation from California (15%, 4½74), Florida (9%, 24/274), Texas (7%, 19/274), and New York (7%, 18/274). Fifty-two percent of respondents categorized the setting of their practice as suburban (146/281), 42% as Urban (117/281), 5% as Rural (15/281) and 1% as other (3/281). With respect to practice setting, 40% work in a “Retina-only practice with greater than 2 retina specialists in the group” (112/280), 20% in a “Multi-specialty ophthalmology group” (57/280), 20% in an “Academic/University medical group” (56/280), 16% in “Solo or less than 3 retinal specialists in a retinal only group” (45/280), and 4% in a “Large multi-specialty medical group” (10/280). Around half of the respondents perform between 30–80 IVI per week (138/281), 20% perform between 0 and 30 IVI per week (57/281), and 31% perform 80 or more IVI per week (86/281).

Pre-Injection Procedures

The results for the pre-injection practices are detailed in Table 1. Briefly, seventy-two percent of respondents reported a nurse or assistant present in the room to help in preparation (20½81). There was a wide array of responses with respect to the type(s) of anesthetics used prior to injection. Among respondents, 16% reported using a topical anesthetic only (46/281), 14% using a topical anesthetic and cotton swab compression (40/281), 31% using an anesthetic gel only or in addition to a topical anesthetic (88/281), and 27% using a subconjunctival anesthetic only or in addition to a topical anesthetic (77/281). We then performed further a multivariate analysis looking specifically at this question of type of anesthetic used. The variables included in the analysis were age, gender, years practiced, geographic location, and practice setting. We did not find a statistically significant association of age, gender, years practiced, or geographic region on retinal specialists’ choice of anesthetic regimen for IVI (p>0.05 for all variables and are available upon request).

The majority of respondents used a lid speculum every time they performed an IVI (67%, 187/281), while 22% never used one (62/281). In terms of applying betadine on the eye, 63% of respondents dropped betadine from a bottle (177/281). If a patient stated that they have a betadine allergy, 58% still used betadine as they believe the allergy did not exist (16½78), while 36% chose a strategy not offered in the choices (100/278). Among those who chose to handle the allergy with alternative means, 22% stated that they would administer antibiotics either before or after the injection (22/100), and 21% would minimize exposure of Betadine to only the injection site or by decreasing the concentration (21/100).

In other aspects, the majority used gloves every time (73%, 145/199) and swabbed eyelids and eyelashes with betadine every time (59%, 117/199). However, only 36% used a mask every procedure (72/199) and only 21% used topical antibiotics prior to injection (42/199). While 45% of respondents dilated the eyes prior to injection (127/281) every procedure, 37% only dilated the eyes half or less than half the time (104/281) prior to injection. The majority of respondents did not perform anterior chamber paracentesis for patients at risk for conditions such as Glaucoma (63%, 177/280).

Injection Procedures

While the majority of respondents used a 30-gauge needle (61%, 170/279), 21% elected to use a 31-gauge needle (59/279) (Table 2). Most respondents injected in the inferior temporal quadrant (70%, 195/279), though 27% chose to inject in the superior temporal quadrant (74/279). If a patient required bilateral injections, 31% stated they would perform it on the same day in every situation (88/281), while 25% would never perform it on the same day (69/281). Most of the retinal specialists who responded did not measure the distance from the limbus upon injecting (69%, 195/281), though 20% utilized calipers (55/281). There was a bimodal distribution on whether to place a cotton tip over the injection site immediately following injection. While 55% never placed a cotton tip immediately over the spot (153/281), 37% of respondents placed a cotton tip over the injection site every time (103/281).

Post-Injection Procedures

Half of the survey respondents utilized Balanced Salt Solution (BSS) to irrigate betadine out of the eyes (50%, 14½81), while 30% chose “Other” means not listed (84/281), and 20% used 3mL 0.9% sodium chloride (56/281) (Table 3). Among those who chose other means, 35% of them used the commercially available “Eye Wash” solution (29/84). Among respondents, 31% stated that they routinely checked IOP (38/123), 28% routinely administered a topical antibiotic (34/123), and 46% routinely performed “Other” procedures post-injection (57/123). Among those that responded “Other,” 21% checked for visual acuity in some manner (counting fingers, hand motion, etc) (12/57).

Discussion

The number of intravitreal injections performed each year continues to increase as the prevalence of retinal vascular diseases such as age-related macular degeneration (AMD), diabetic retinopathy, and retinal occlusion continue to grow. While a number of previous surveys of IVI practice have been performed, to the best of our knowledge, this is the largest survey that has been performed that details the specific anesthetic agents, antibiotic usage, and various other techniques employed while performing IVI. In particular, in the United States, the last survey was conducted in 2011 by Green-Simms et al.² A more recent survey detailing IVI techniques was conducted in Israel in 2016 as well.⁸ We also reviewed the results from the last five years of the PAT survey from the American Society of Retina Specialists (2014–2018), and only 6 questions over the entire 5 years overlapped with questions in the survey . Though these surveys have assessed important information about best practices for IVI, they do not detail important information regarding specific anesthetics used, betadine allergies, nor assess differences in techniques based on demographic parameters. In this study we performed additional multivariate analysis to assess the effect of demographic variables on anesthetic choice for IVI, and did not find that age, gender, practice type, or geographic location influenced anesthetic choice for IVI.

Pre-Injection Methodology

The study found that the majority of respondents had a nurse or assistant present in the room to help prepare for the injections, which can lead to increased efficiency and decreased errors in preparation.⁹ Interestingly, anesthetic choice did not differ based on gender, age, practice setting, or geographic location. Previous studies have not demonstrated a significant difference in pain score between the different types of anesthetic used for IVI^10,11.

The majority of respondents utilized a lid speculum, which can help to avoid lid reflex closure and subsequent exposure of microbes to the injection site. For those respondents that never utilized a lid speculum or use it less often, there are alternative approaches available such as the “Lid Splinting Approach” that may be helpful in decreasing contamination.¹² Many respondents elected to change their protocol if a patient reported a betadine allergy, even though previous studies have demonstrated increased risk for developing more serious conditions such as endophthalmitis when electing not to use betadine.¹³

There is no consensus on whether to use, drapes, betadine to swab eyelids and eyelashes, gloves, and masks during every IVI. Among the respondents who use at least one of these items every procedure, 21% administered topical antibiotics, despite the increasing body of literature that shows the rates of endophthalmitis being similar when using antibiotics drops prophylactically as compared to control.¹⁴ Other studies have demonstrated the possibility of there being inherent harm to utilizing antibiotics pre-injection through increasing resistance to bacterial organisms.^15,16 There was a minority of respondents who reported regularly using masks, but there are studies that show significant associations between speech and post-injection infection rates in the setting of intravitreous injections.^17,18 Though not many respondents performed AC paracentesis in patients with glaucoma, previous studies indicate there may be benefit in performing AC paracentesis in high-risk individuals with glaucoma and or cotton tip decompression.^19–21

Injection Methodology

Some previous studies have shown no significant difference between using 27 and 30-gauge needles in terms of patient pain scores.²² In contrast, alternate studies find that there is a significant decrease in reported pain when using a 30-gauge needle versus a 27-gauge.²³ Though there is conflicting evidence as to whether the 30-gauge needle leads to less pain, physicians prefer using the 30-gauge, which is also corroborated by this study. There does not seem to be a difference in patients’ reported pain scores when comparing 30 and 31 gauge needles, and 33 to 30 gauge needles.^24,25

Over a quarter of respondents chose to inject in the superior temporal quadrant; however, recent studies have shown that the superior temporal quadrant is the most painful for patients and should be avoided if possible.²⁶ While those that performed more IVI per week tended to perform more bilateral injections on the same day, the majority elect not to. Though it is possible to perform bilateral IVI safely, studies have shown that each eye should be treated as a separate procedure to decrease complication rates.²⁷ While most respondents did not measure the distance from the limbus, previous studies have shown that injecting posterior to 4.5 mm from the limbus can lead to retinal tears, and thus beginning residents and fellows may find it useful to measure the distance with a caliper or alternate tool at least in the beginning stages of their practice.^8,28

Post-Injection Methodology

Though most respondents never placed a cotton tip over the injection site following IVI, previous guidelines recommend the placement of a cotton tip on the injection site to prevent possible reflux.²⁹ The majority of respondents did not perform a dilated fundus exam. Though there are not many studies documenting the association between a dilated exam and specific outcomes, it can be beneficial in terms of viewing the retinal artery perfusion post injection.³⁰ With regard to irrigating betadine out of the eye, previous studies have shown that Povidone-iodine (PI) is associated with greater discomfort as compared to other antiseptic agents such as aqueous chlorhexidine.³¹

Though most respondents did not administer topical NSAIDs or steroids after injection, there is some evidence that NSAIDs post injection for macular edema can lead to mild decreases in central retinal thickness and may reduce the number of anti-VEGF therapies administered in patients who require IVI frequently.³² NSAIDs have also been shown to reduce pain at 6 hours and 24 hours after the administration of IVI.³³ NSAIDs can thus be considered as a post injection measure, especially for patients who experience higher than normal pain levels with the procedure. One point of debate in the community is whether to use antibiotics post-injection. Previous studies demonstrate that post IVI antibiotic administration can lead to increased resistance to virulent organisms such as coagulase negative Staphylococcus.³⁴ Furthermore, checking IOP or assessing for light perception can be an important way to ensure that the central retinal artery remains perfused.³⁰

There are multiple limitations to this study. The response rate was relatively low (17% of those surveyed). While we attempted to recruit a large sample size, response bias and other factors could skew the results of the survey. While it would be interesting to look at multivariate analysis for multiple questions for this survey, we decided to only look at type of anesthetic used, to avoid false positives from multi-hypothesis testing. Additionally, this survey was sent to a large subset of retina specialists in the United Stateswhich could introduce bias in responses. In conclusion, we believe this study provides an updated survey of current practices of retina specialists in the United States, and provides critical information regarding current real world practices for intravitreal injection administration.

Abbreviations:

IVI

Intravitreal Injections

Appendix

Survey

Demographic information

1. What is your age?

2. To which gender identity do you most identify with?

   • Female
   • Male
   • Prefer not to answer

3. How many years have you been practicing post training in the field of ophthalmology?

4. In which state do you practice medicine in the field of ophthalmology?

5. Which of the following practice settings do you work in?

   • Multispecialty ophthalmology group
   • Large multi-specialty medical group (nonacademic, e.g. Kaiser)
   • Academic/university medical group
   • Retina only practice (>2 retina specialists in group)
   • Solo or less than 3 retina specialists in a retina only group
6. How would you describe the setting of the practice that you work in?

   • Urban
   • Suburban
   • Rural
   • Other: please explain

Pre-Injection

1. Is a nurse or assistant present in the room to help prepare the patient for the injection?

   • Yes
   • No
   • Sometimes
2. Please select the type(s) of anesthesia used prior to injection:

   • topical proparacaine
   • topical proparacaine plus cotton swab compression
   • lidocaine gel
   • subconjunctival lidocaine
   • other (please explain):
3. What percentage of the time do you use a lid speculum?

   • 0%
   • 25%
   • 75%
   • 100%
4. How do you apply betadine on the eye?

   • Cotton tip applicator
   • Drop from a betadine bottle
   • Use a TB syringe
   • Other (please explain):
5. What do you do if a patient states they have a “betadine allergy?”

   • I still use betadine as there is no such thing as a betadine allergy
   • Use chlorhexidine
   • Refuse to inject them and refer patient to somewhere else
   • Other (please explain):
6. Please select all materials used during every procedure:

   • Topical antibiotic prior to injection
   • Drape
   • Swab eyelids and eyelashes with betadine
   • Gloves
   • Mask
7. What percentage of the time do you dilate the eyes prior to injection?

   • 0%
   • 25%
   • 50%
   • 75%
   • 100%
8. What percentage of the time do you perform AC paracentesis for patients at risk for damage from elevated intraocular pressures (e.g. glaucoma).

   • 0%
   • 25%
   • 50%
   • 75%
   • 100%
   • I perform ocular decompression with a cotton tip
   • Other (please explain):
9. How many intravitreal injections do you perform per week?

   • 0–5
   • 5–30
   • 30–80
   • 80+

Injection

1. Which of the following needle gauges do you typically use for anti-VEGF injections?

   • 27
   • 30
   • 31
   • 33
   • Other (please explain):
2. Which quadrant do you typically inject in?

   • Superior temporal quadrant
   • Inferior temporal quadrant
   • Superior nasal quadrant
   • Inferior nasal quadrant
3. What percentage of the time do you perform bilateral injections the same day if the patient requires injections in both eyes?

   • 0%
   • 25%
   • 50%
   • 75%
   • 100%
4. How do you measure the distance from the limbus?

   • Calipers
   • I use another device to measure the distance (e.g. Tb syringe)
   • I do not measure the distance and approximate instead
   • Other (please explain):
5. What percentage of the time do you place a cotton tip over the site immediately following injection?

   • 0%
   • 25%
   • 50%
   • 75%
   • 100%
6. What percentage of the time do you do a dilated exam when doing injections?

   • 0%
   • 25%
   • 50%
   • 75%
   • 100%

Post-Injection

1. What do you use to irrigate betadine out of the eye?

   • 3 mL 0.9% sodium chloride
   • BSS
   • Other (Please explain):
2. Do you routinely use any of the following after injection?

   • Topical NSAIDS/steroids after injection
   • Topical antibiotics
   • Check IOP
   • Other (please explain):

3. Is there anything else you would like to add about your intravitreal injection practice that was not covered?

Table 1.

Pre-injection Practices of American Retina Specialists

Question	Responses	n	%
Nurse or Assistant Present	Always Present	201 / 281	72
	Sometimes Present	26 / 281	9
	Never Present	54 / 281	19
Anesthetic Choice	Topical Anesthetic Only	46 / 281	16
	Topical Anesthetic and Cotton Swab Compression	40 / 281	14
	Anesthetic Gel Only or Anesthetic Gel and Topical Anesthetic Combination	88 / 281	31
	Subconjunctival Anesthetic Only or Subconjunctival and Topical Anesthetic Combination	77 / 281	27
	Alternate Combination Not Listed	30 / 281	11
Lid Speculum Usage	Always Use	187 / 281	67
	Use 75% of the Time	16 / 281	6
	Use 25% of the Time	16 / 281	6
	Never Use	62 / 281	22
Betadine Application	Cotton Tip Applicator	50 / 281	18
	Drop from a Betadine Bottle	177 / 281	63
	TB Syringe	21 / 281	8
	Alternate Method	33 / 281	12
Betadine Allergy Response to Patient with Allergy	Still Use Betadine since an Allergy Does Not Exist	161 / 278	58
	Use Chlorhexidine	15 / 278	5
	Refuse Injection and Refer to Different Location	2 / 278	1
	Alternate Method	100 / 278	36
Materials Used Every Procedure	Topical Antibiotics	42 / 199	21
	Drape	13 / 199	7
	Swab Eyelids and Eyelashes with Betadine	117 / 199	59
	Gloves	145 / 199	73
	Mask	72 / 199	36
Dilation of Eyes Prior to Injection	Always Dilate	127 / 281	45
	Dilate 75% of the Time	50 / 281	18
	Dilate 50% of the Time	40 / 281	14
	Dilate 25% of the Time	41 /281	15
	Never Dilate	23 / 281	8
AC Paracentesis For At- Risk Patients	Never Perform	177 / 281	63
	Perform 25% of the Time	28 / 281	10

Table 2.

Injection Practices of American Retina Specialists

Question	Responses	n	%
Needle Gauge Preference	27 Gauge	3 / 279	1
	30 Gauge	170 / 279	61
	31 Gauge	59 / 279	21
	33 Gauge	18 / 279	7
	Other	29 / 279	10
Injection Quadrant	Superior Temporal	74 / 279	27
	Inferior Temporal	195 / 279	70
	Superior Nasal	0 / 279	0
	Inferior Nasal	10 / 279	4
Percentage of Time Physician Performs Bilateral Injections if Required	0%	69 / 281	25
	25%	34 / 281	12
	50%	19 / 281	7
	75%	71 / 281	25
	100%	88 / 281	31
Method for Measuring Distance From Limbus	Calipers	55 / 281	20
	Alternate Device (e.g. Tb syringe)	29 / 281	10
	Do Not Measure	195 / 281	69
	Other	2 / 281	1
Percentage of Time Physician Places Cotton Tip over Injection Site	0%	153 / 281	55
	25%	15 / 281	5
	50%	1 / 281	<1
	75%	9 / 281	3
	100%	103 / 281	37
Percentage of Time Physician Performs Dilated Exam	0%	33 / 281	12
	25%	65 / 281	23
	50%	42 / 281	15
	75%	53 / 281	19
	100%	88 / 281	31

Table 3.

Post-Injection Practices of American Retina Specialists

Question	Responses	n	%
Irrigation Method for Betadine	3 mL 0.9% Sodium Chloride	56 / 281	20
	Balanced Salt Solution (BSS)	141 / 281	50
	Other	84 / 281	30
Routinely Used Materials and Methods	Topical NSAIDs/Steroids	17 / 123	14
	Topical Antibiotics	34 / 123	28
	Check IOP	38 / 123	31
	Other	57 / 123	46