Patterns and Disparities in Recorded Gonioscopy During Initial Glaucoma Evaluations in the United States

Abstract

PURPOSE:

To assess patterns in gonioscopy during initial glaucoma evaluations in the United States.

DESIGN:

Retrospective, case-control study.

METHODS:

Patients undergoing initial glaucoma evaluation between 2009–2020 were identified in the Optum Clinformatics^® DataMart. Initial evaluation was defined as: 1) glaucoma suspect, anatomical narrow angle (ANA), or primary/secondary glaucoma diagnosed by an ophthalmologist; 2) continuously observable during a 36-month lookback period; 3) no history of glaucoma medications, laser, or surgical procedures; 4) OCT or visual field performed within 6 months of initial diagnosis. Logistic regression models were developed to identify factors associated with no record of gonioscopy based on Current Procedural Terminology (CPT) codes.

RESULTS:

Among 198,995 patients, 20.4% and 29.5% had recorded gonioscopy on the day of diagnosis or within 6 months, respectively. On multivariable analysis, odds of recorded gonioscopy within 6 months of initial evaluation was lower (p<0.001) among non-Hispanic Whites (OR=0.84) but similar for Blacks (OR=1.02) and Hispanics (OR=0.96) compared to Asians. Age over 60 years (OR<0.82), pseudophakia/aphakia (OR=0.58), or residence outside of the Northeast region (OR=0.66–0.84) conferred lower odds of recorded gonioscopy (p<0.001). Angle closure glaucoma (OR=0.85), secondary glaucoma (OR=0.31), or open angle glaucoma/suspect (OR=0.12/0.24, respectively) patients were less likely to have recorded gonioscopy compared to ANA patients (p<0.01).

CONCLUSIONS:

Over 70% patients undergoing initial glaucoma evaluation in the United States do not have record of gonioscopy, especially elderly, non-Hispanic White, and pseudophakic patients in non-Northeast regions. This pattern does not conform to current practice guidelines and could contribute to misdiagnosed disease and suboptimal outcomes.

Graphical Abstract

The rate of gonioscopy during initial glaucoma evaluation between 2009–2020 in United States was assessed using healthcare claims data. The majority of patients undergoing initial glaucoma evaluation do not have record of gonioscopy. There are also racial and other sociodemographic disparities in gonioscopy rates. The observed patterns in gonioscopy do not conform to practice guidelines and may contribute to glaucoma misdiagnosis and ocular morbidity.

Introduction

Glaucoma is a leading cause of blindness in the world.¹ The two primary types of glaucoma are primary open angle glaucoma (POAG) and primary angle closure glaucoma (PACG). Although POAG is more common than PACG in the United States, PACG is associated with more severe ocular morbidity and confers three-fold higher risk for severe bilateral visual impairment and unilateral blindness compared to POAG.^2–5 Therefore, early and accurate detection of patients with or at risk for PACG is crucial, especially since treatment guidelines for PACG and POAG differ and treatment with laser peripheral iridotomy (LPI) and lens extraction can effectively alleviate angle closure and prevent glaucomatous damage to the optic nerve.⁶ Conversely, the misclassification of glaucoma subtypes during initial glaucoma evaluation may lead to missed opportunities for intervention and increased visual morbidity.^7,8

Gonioscopy is the current clinical standard for detecting patients at risk for PACG and plays a key role in the comprehensive glaucoma evaluation.⁹ The American Academy of Ophthalmology (AAO) and World Glaucoma Association (WGA) both recommend that all patients receive gonioscopy at initial evaluation for glaucoma and every five years thereafter.^10,11 Despite the clinical significance and emphasis placed on gonioscopy by these guidelines, there is evidence that gonioscopy is underperformed by eyecare providers in the United States. Previous studies showed that gonioscopy rates are highly variable based on clinical setting, and less than half of patients undergoing glaucoma surgery had record of gonioscopy in the preceding five years.^12–15 However, there is a sparse knowledge about specific risk factors associated with lack of gonioscopy at initial glaucoma evaluation in a large, diverse cohort.

In this study, we use national healthcare claims data to study patients undergoing initial glaucoma evaluations and assess the proportion of patients who had recorded gonioscopy during these encounters. Our aim is also to identify sociodemographic and clinical characteristics associated with lower likelihood of recorded gonioscopy during the initial glaucoma evaluation period. We hypothesize some patient populations are less likely to receive gonioscopy, which could contribute to later detection of angle closure, misdiagnosis of glaucoma type, and higher risk of visual morbidity.

Methods

Data

Optum’s Clinformatics^® Data Mart (CDM) is derived from a database of administrative health claims data warehouse of commercial and Medicare Advantage health claims. The database includes approximately 17 to 19 million annual covered lives, for a total of over 65 million unique lives over a 12-year period (1/2009 through 12/2020). Available clinical data included glaucoma subtype diagnosis with first date of diagnosis, healthcare provider type, in-office procedures including gonioscopy, optical coherence tomography (OCT) retinal nerve fiber layer (RNFL) or ganglion cell complex (GCC), visual field testing, cataract and glaucoma surgeries, lens status (cataract, pseudophakia/aphakia/history of lens extraction, or no lens diagnosis), and first/last date of enrollment. Available sociodemographic data included sex, race, education level, net worth, insurance product, and census division regions. The University of Southern California Institutional Review Board determined that this study was exempt from IRB approval. The study adhered to the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act.

Study Population and Definitions

Inclusion in the study population required an index date of diagnosis based on International Classification of Diseases, Ninth Revision (ICD9) or Tenth Revision (ICD10) codes (Supplementary Table 1). The index date of diagnosis was defined as the date of the first diagnosis of any glaucoma or glaucoma suspect. Inclusion also required continuous enrollment and observability, calculated as the number of months between the first date of enrollment minus the index date of initial glaucoma diagnosis, for at least 36 months prior to the index date of diagnosis.

Initial evaluation of glaucoma met the following criteria: (1) glaucoma suspect, anatomical narrow angle, or glaucoma (open angle, angle closure, or secondary) diagnosis provided by an ophthalmologist at age 18 or older, (2) OCT RNFL or GCC or visual field testing performed on or within 6 months of the index date of diagnosis, (3) continuous enrollment and observability during a 36-month lookback period, (4) no history of drops, laser procedures, or glaucoma surgery based on Current Procedural Terminology (CPT) codes (Supplementary Table 1) before the index glaucoma diagnosis date. Criteria 1 and 2 were implemented to increase the fidelity of glaucoma diagnosis codes. Criterion 3 was implemented to minimize mislabeling of established evaluations as initial evaluations. Criterion 4 was implemented to ensure patients who received conventional glaucoma treatments without the diagnosis glaucoma would not be included in the cohort of initial evaluations. Record of gonioscopy during initial glaucoma evaluation, the primary outcome measure, was based on CPT codes within the 6-month period after the index diagnosis of glaucoma. Patients with multiple glaucoma diagnosis codes were classified based on the more severe diagnosis (glaucoma over glaucoma suspect).

Statistical Analysis

The proportion of initial glaucoma evaluations with recorded gonioscopy within 6 months was stratified by sex, age, race, and lens status. Analyses were conducted on the patient level rather than eye level due to lack of laterality data in ICD9. Continuous data were expressed as means and standard deviations while categorical data were expressed as proportions and percentages. Univariable logistic regression models were developed with the sociodemographic variables, glaucoma subtype diagnosis, lens status, and observable time to determine odds ratios (OR) for recorded gonioscopy. A multivariable regression model was developed with age, sex, and variables significant at P < 0.15 in the univariable model. From this multivariable model, clinically and statistically significant variables were defined as having an OR greater than 1.15 or less than 0.85 and a P value < 0.01. A secondary analysis was conducted to assess the proportion of initial glaucoma evaluations receiving gonioscopy on the same day as the initial evaluation. Statistical analysis was completed with R version 4.2.1.

Results

A total of 198,995 patients undergoing initial glaucoma evaluation and meeting all inclusion criteria were identified in the Optum database. About 56.0% of the cohort were female, and there were 65.1% non-Hispanic Whites, 6.1% Asians, 12.9% Blacks, 12.1% Hispanics and 3.8% of unknown race. The average age at diagnosis was 66.0±14.1 years (range, 18 to 90 years). 2.1% of patients had more than one glaucoma diagnosis code. After selection of the more severe diagnosis (glaucoma over glaucoma suspect), 81.1% were diagnosed as open angle glaucoma suspect (OAGS), 5.7% were diagnosed with anatomical narrow angle (ANA), 10.1% were diagnosed with POAG, 1.2% were diagnosed with PACG, and 1.9% were diagnosed with secondary glaucoma (SG). 41.2% were coded as phakic with cataract and 10.6% were coded as pseudophakic or aphakic or had history of cataract extraction during the lookback period. The remaining 48.2% had no coded lens status.

The proportion of patients with recorded gonioscopy was 20.4% on the same day of diagnosis and 29.5% within 6 months of index date of diagnosis across all ages, sexes, and races. The proportion of male and female cases with recorded gonioscopy within 6 months was similar (29.7% and 29.4%, respectively; Table 1). The racial group with the highest proportion of recorded gonioscopy within 6 months (Table 2) was, in order, Asians (35.3%), Hispanics (32.7%), Blacks (32.6%), and non-Hispanic Whites (27.9%). Among patients with PACG, only 56.6% of Asians had record of gonioscopy compared to 73.5% of Blacks, 70.4% of non-Hispanic Whites, and 70.3% of Hispanics. Other glaucoma subtypes had less variability between races, with 24.6% of OAGS (range between races, 23.1% - 30.6%), 72.6% of ANA (range between races, 71.7% - 74.2%), 37.3% of POAG (range between races, 35.3% - 43.0%), and 42.1% of SG (range between races, 41.6% - 45.1%) having recorded gonioscopy within 6 months of diagnosis (Table 3). About 30.6% of those with cataracts had recorded gonioscopy within 6 months compared to 17.4% of patients with pseudophakia, aphakia, or history of cataract extraction in the previous 36 months and 31.4% of patients with no lens status diagnosis (Table 4).

Table 1.

Proportion of all initial glaucoma evaluations receiving gonioscopy within 6 months stratified by age and sex.

Age Range	Male			Female			Overall
	N	n	%	N	n	%	N	n	%
<40	4866	1612	33.13%	5166	1608	31.13%	10034	3221	32.10%
40–49	7574	2661	35.13%	8973	3200	35.66%	16547	5861	35.42%
50–59	13674	4523	33.08%	17564	6110	34.79%	31248	10637	34.04%
60–69	18881	5822	30.84%	24223	7448	30.75%	43124	13278	30.79%
70–79	28936	7993	27.62%	36920	10051	27.22%	65866	18050	27.40%
80+	13501	3385	25.07%	18666	4360	23.36%	32176	7745	24.07%
Overall	87432	25996	29.73%	111512	32777	29.39%	198995	58792	29.54%

Abbreviations: N = Initial glaucoma evaluation cases; n = Initial glaucoma evaluation cases receiving gonioscopy within 6 months.

Table 2.

Proportion of all initial glaucoma evaluations receiving gonioscopy within 6 months stratified by age and race.

Age Range	Asian			Black			Hispanic			Non-Hispanic White
	N	n	%	N	n	%	N	n	%	N	n	%
<40	1412	490	34.70%	1311	408	31.12%	1648	573	34.77%	5269	1624	30.82%
40–49	1881	705	37.48%	2385	866	36.31%	2712	986	36.36%	8973	3094	34.48%
50–59	2044	763	37.33%	4600	1616	35.13%	4047	1483	36.64%	19479	6421	32.96%
60–69	2115	757	35.79%	5846	1981	33.89%	4627	1556	33.63%	28937	8512	29.42%
70–79	3390	1188	35.04%	8053	2486	30.87%	7747	2374	30.64%	43823	11274	25.73%
80+	1256	367	29.22%	3385	990	29.25%	3349	918	27.41%	23018	5228	22.71%
Overall	12098	4270	35.30%	25580	8347	32.63%	24130	7890	32.70%	129499	36153	27.92%

Abbreviations: N = Initial glaucoma evaluation cases; n = Initial glaucoma evaluation cases receiving gonioscopy within 6 months.

Table 3.

Proportion of all initial glaucoma evaluations receiving gonioscopy within 6 months stratified by glaucoma diagnosis and race.

Glaucoma Subtype	Asian			Black			Hispanic			Non-Hispanic White			Overall
	N	N	%	N	n	%	N	n	%	N	n	%	N	n	%
Open Angle Glaucoma Suspect	9770	2988	30.58%	20249	5658	27.94%	19078	5201	27.26%	105938	24414	23.05%	161330	39757	24.64%
Angle Closure Without Glaucoma	1034	741	71.66%	973	722	74.20%	1742	1250	71.76%	7239	5273	72.84%	11416	8293	72.64%
Open Angle Glaucoma	964	367	38.07%	3711	1596	43.01%	2613	1045	39.99%	12052	4259	35.34%	20087	7498	37.33%
Angle Closure Glaucoma	221	125	56.56%	279	205	73.48%	320	225	70.31%	1500	1056	70.40%	2417	1674	69.26%
Secondary Glaucoma	109	49	44.95%	368	166	45.11%	377	169	44.83%	2770	1151	41.55%	3745	1577	42.11%
Overall	12098	4270	35.30%	25580	8347	32.63%	24130	7890	32.70%	129499	36153	27.92%	198995	58799	29.55%

Abbreviations: N = Initial glaucoma evaluation cases; n = Initial glaucoma evaluation cases receiving gonioscopy within 6 months.

Table 4.

Proportion of all initial glaucoma evaluations receiving gonioscopy within 6 months stratified by age and lens status.

Age Range	No Lens Diagnosis			Pseudophakia, aphakia, or history of cataract extractiona			Cataract
	N	n	%	N	n	%	N	n	%
<40	9662	3087	31.95%	50	12	24.00%	322	122	37.89%
40–49	14409	5101	35.40%	185	47	25.41%	1953	713	36.51%
50–59	20976	7201	34.33%	863	177	20.51%	9409	3260	34.65%
60–69	18426	6006	32.60%	2974	539	18.12%	21724	6734	31.00%
70–79	20152	5803	28.80%	9485	1631	17.20%	36229	10616	29.30%
80+	12326	2882	23.38%	7562	1259	16.65%	12288	3609	29.37%
Overall	95951	30080	31.35%	21119	3665	17.35%	81925	25054	30.58%

Abbreviations: N = Initial glaucoma evaluation cases; n = Initial glaucoma evaluation cases receiving gonioscopy within 6 months.

^aMore remote than 36-month lookback.

On multivariable analysis (Table 5), patients 60 years and older were less likely to have recorded gonioscopy within 6 months than their younger counterparts (OR ≤ 0.82, p < 0.001). Non-Hispanic Whites were less likely than Asians to receive gonioscopy overall (OR = 0.84, p < 0.001), and those residing outside of the Northeast were significantly less likely to receive gonioscopy (OR ≤ 0.84, p < 0.001) than those living in the Northeast. Patients with diagnoses of pseudophakia, aphakia, or history of cataract extraction during the lookback period were less likely to receive gonioscopy than those diagnosed with cataract (OR = 0.58, p < 0.001). The OAGS group (OR = 0.12, p < 0.001), the POAG group (OR = 0.24, p < 0.001), the SG group (OR = 0.31, p < 0.001), and the PACG group (OR = 0.85, p = 0.001) were less likely to have recorded gonioscopy within 6 months of initial evaluation compared to the ANA group. A secondary analysis was performed to assess the rate of gonioscopy performed on the same day as initial glaucoma evaluation rather than within 6 months of the evaluation. Similar results were observed except the difference between non-Hispanic Whites and Asians approached but did not reach the threshold for clinical significance (OR = 0.87, p < 0.001). There was also an additional effect of those with exclusive provider organization (EPO) insurance being more likely to have recorded gonioscopy to those with health maintenance organization (HMO) insurance (OR = 1.21, p < 0.001).

Table 5.

Univariable and multivariable analysis of factors associated with receiving gonioscopy within 6 months.

		Gonioscopy received	Univariable Analysis		Multivariable Analysis
		No. (%)	OR (95% CI)	P value	OR (95% CI)	P value
Age (years)
	<40	3221 (32.1)	0.86 (0.82–0.91)	<0.001	0.91 (0.86–0.96)	0.001
	40–49	5861 (35.4)	REF		REF
	50–59	10638 (34.0)	0.94 (0.90–0.98)	0.003	0.94 (0.90–0.98)	0.002
	60–69	13279 (30.8)	0.81 (0.78–0.84)	<0.001	0.82 (0.79–0.85)	<0.001
	70–79	18050 (27.4)	0.69 (0.66–0.71)	<0.001	0.69 (0.67–0.72)	<0.001
	80+	7750 (24.1)	0.58 (0.56–0.60)	<0.001	0.62 (0.59–0.64)	<0.001
Sex a
	Female	32777 (29.4)	REF		REF
	Male	25996 (29.7)	1.02 (1.00–1.04)	0.099	1.05 (1.03–1.07)	<0.001
Race
	Asian	4270 (35.3)	REF		REF
	Non-Hispanic White	36153 (27.9)	0.71 (0.68–0.74)	<0.001	0.84 (0.81–0.88)	<0.001
	Black	8347 (32.6)	0.89 (0.85–0.93)	<0.001	1.02 (0.97–1.07)	0.497
	Hispanic	7890 (32.7)	0.89 (0.85–0.93)	<0.001	0.96 (0.91–1.01)	0.114
	Unknown	2139 (27.8)	0.71 (0.66–0.75)	<0.001	0.85 (0.79–0.93)	<0.001
Education
	H.S. Diploma	13242 (29.9)	REF		REF
	Some College	30478 (29.1)	0.96 (0.94–0.98)	0.001	1.05 (1.02–1.08)	0.001
	College Plus	13692 (30.4)	1.02 (0.99–1.05)	0.170	1.09 (1.05–1.13)	<0.001
	Unknown	1084 (26.8)	0.85 (0.79–0.92)	<0.001	0.96 (0.71–1.27)	0.761
Net Worth
	<$149k	18100 (29.9)	REF		REF
	$150k– $499k	15452 (29.4)	0.97 (0.95–1.00)	0.040	1.00 (0.97–1.03)	0.910
	> $500k	21461 (29.2)	0.97 (0.94–0.99)	0.005	0.97 (0.94–1.00)	0.056
	Unknown	2799 (32.2)	1.11 (1.06–1.17)	<0.001	1.06 (1.01–1.12)	0.031
Insurance
	HMO	10299 (26.1)	REF		REF
	EPO	4524 (35.4)	1.55 (1.49–1.62)	<0.001	1.13 (1.07–1.18)	<0.001
	PPO	4425 (30.3)	1.23 (1.18–1.28)	<0.001	1.06 (1.01–1.11)	0.011
	Other	39551 (29.9)	1.21 (1.18–1.24)	<0.001	1.07 (1.04–1.10)	<0.001
Location
	Northeast	11077 (36.0)	REF		REF
	Mountain	3808 (23.9)	0.56 (0.53–0.58)	<0.001	0.67 (0.64–0.70)	<0.001
	Midwest	9315 (24.6)	0.58 (0.56–0.60)	<0.001	0.66 (0.64–0.68)	<0.001
	Pacific	5651 (27.7)	0.68 (0.66–0.71)	<0.001	0.75 (0.72–0.78)	<0.001
	South	28825 (30.8)	0.79 (0.77–0.81)	<0.001	0.84 (0.82–0.87)	<0.001
Glaucoma Subtype
	Open Angle Glaucoma Suspect	39757 (24.6)	0.12 (0.12–0.13)	<0.001	0.12 (0.12–0.13)	<0.001
	Angle Closure Without Glaucoma	8293 (72.6)	REF		REF
	Open Angle Glaucoma	7498 (37.3)	0.22 (0.21–0.24)	<0.001	0.24 (0.23–0.25)	<0.001
	Angle Closure Glaucoma	1674 (69.3)	0.85 (0.77–0.93)	<0.001	0.85 (0.77–0.93)	0.001
	Secondary Glaucoma	1577 (42.1)	0.27 (0.25–0.30)	<0.001	0.31 (0.29–0.33)	<0.001
Lens
	Cataract	25054 (30.6)	REF		REF
	No lens diagnosis	30080 (31.3)	1.04 (1.02–1.06)	<0.001	0.97 (0.95–0.99)	0.006
	Pseudophakia, aphakia, or history of cataract extractionb	3665 (17.4)	0.48 (0.46–0.50)	<0.001	0.58 (0.56–0.61)	<0.001
Observable period (years)
	Mean (SD)	3.0 (2.4)	1.00 (1.00–1.01)	0.028	1.02 (1.01–1.02)	<0.001

Statistically significant p-values and odds ratios are bolded. Abbreviations: CI = Confidence Interval; EPO = exclusive provider organization; HMO = health maintenance organization; NA = not applicable; OR = odds ratio; PPO = preferred provider organization; REF = reference group; SD = standard deviation.

^aSex was unknown for 61 subjects.

^bMore remote than 36-month lookback.

Discussion

Overall, only around 30% of patients undergoing initial glaucoma evaluation had recorded gonioscopy despite clear recommendations by the AAO and WGA. The proportion with recorded gonioscopy was significantly lower among non-Hispanic White patients compared to racial minorities, patients 60 years of age and older, patients with diagnosis of pseudophakia/aphakia, and/or residing outside of the Northeast geographic region. Patients diagnosed as ANA were more likely to receive gonioscopy compared to patients with other glaucoma diagnoses. Our observed patterns in recorded gonioscopy do not conform to current practice guidelines, highlighting the potential for suboptimal care of misdiagnosed glaucoma and need for more convenient angle assessment methods.

The overall low rate of gonioscopy is striking. Only 20.4% and 29.5% of patients were coded to have received gonioscopy on the same day and within 6 months of initial evaluation, respectively. A previous study by Hertzog et al. reported that 51.3% of the patients undergoing initial glaucoma evaluations in a private, community-based ophthalmology practice setting received gonioscopy.¹⁴ Similarly, Coleman et al., who used the data from a random 5% sample of Medicare beneficiaries, reported a recorded gonioscopy rate of 46% in the 5 years prior to glaucoma surgery being performed in the United States.¹² The difference between our results and those by Hertzog et al., who conducted manual review of clinic notes, could be partially explained by undercoding. However, in the present study, even the proportion of patients with diagnosis of ANA or PACG and recorded gonioscopy within 6 months was only 72.1%, suggesting that a significant number of cases of angle closure were either assumed based on slit lamp exam alone or received gonioscopy that was not coded. While it cannot be assumed that providers are aware of the reimbursement eligibility of gonioscopy, patterns among medical codes within ophthalmology are often complex and their accuracy varies by test, procedure, or provider type.^16,17 As for Coleman et al., the significant difference in the study population (patients receiving glaucoma surgery versus those undergoing initial evaluation) may have contributed to the divergence in findings.

Overall, our present findings reiterate that gonioscopy is likely underperformed and/or underbilled on a nationwide level despite its clinical importance. Gonioscopy represents a crucial junction in the glaucoma management algorithm where appropriate therapy could be prescribed to prevent permanent morbidity. In a recent study, Wu et al. reported that while a majority (67.5%) of Medicare beneficiaries diagnosed with acute angle closure glaucoma received an ocular examination within the 2 years preceding the angle closure, less than a third of them (33.2%) had record of gonioscopy.¹⁸ Our findings support their conclusion that an immense number of opportunities to avoid complications related to angle closure are missed and the need for additional research identifying high-risk individuals.^19–23

According to our multivariable model, Asians, Hispanics, and Blacks were more likely to have recorded gonioscopy compared to non-Hispanic Whites, although the rates were still low overall. Many studies have highlighted the higher burden of angle closure in Asian populations.^6,24–26 Therefore, it is intuitive that the Asian patients in the present study had the highest rates of recorded gonioscopy (35.3%) overall. The similar overall rates of recorded gonioscopy between Asians, Hispanics, and Blacks, however, is interesting. Primary factors that influence a provider’s decision to perform gonioscopy during glaucoma evaluation include perceived risk of angle closure based on patient demographics (e.g. Asian race or older age) or clinical findings associated with angle closure (e.g. shallow anterior chamber depth or hyperopic refractive error).²⁷ If the former was predominant, we should have observed significantly higher rates of recorded gonioscopy among Asians, which we did not. One interpretation of the similar rates of recorded gonioscopy among Asians, Blacks, and Hispanics is that angle closure is more common among Blacks and Hispanics than previously reported. However, it is difficult to reconcile these similar rates of recorded gonioscopy with recent studies reporting higher risk of blindness and need for glaucoma surgery among Blacks and Hispanics with PACG than non-Hispanic Whites.^8,28. This discrepancy appears to indicate that suboptimal access to care - well documented in many facets of medical care for racial minorities – rather than quality of care received may underlie the disproportionate burden of PACG in these racial groups.

When data were stratified by race and glaucoma type, Asian patients with PACG received less gonioscopy (56.6%) compared to other patients with PACG diagnosis (>69.3%). A coding deficiency that affects this specific group seems unlikely. Cognitive bias, or more specifically anchoring, is a shortcut taken by the brain to expedite medical decisions.²⁹ As it is well known that PACG prevalence is highest amongst Asians, providers may be biased towards diagnosing PACG in an Asian patient without confirmatory gonioscopy compared to patients of other races. The providers may have relied on elements of the medical history or other exam findings to arrive at the PACG diagnosis. However, indirect assessment of the angle through anterior chamber appearance, such as by the Van Herick technique, is a poor substitute for gonioscopy. Furthermore, the literature is conflicted on the topic of racial differences in anterior segment biometrics; some studies reported Asian patients have shallower anterior chamber depth while other studies found no difference in anterior chamber depth and axial length between races.^30–32 These findings highlight the need for additional multiracial studies identifying racial differences in biometric risk factors for angle closure that can help elucidate racial difference in recorded gonioscopy rates and angle closure outcomes.

Older age and female sex are well-known risk factors for angle closure and PACG.³³ The increased risk conferred by these two factors is secondary to progressive narrowing of the anterior chamber angle associated with aging and biometric differences between male and female eyes.³⁴ Our study observed that older patients above the age of 60 received less gonioscopy, and there was no significant difference between the sexes. While we expected that the rate of gonioscopy may drop in the older population due to the higher prevalence of pseudophakes, both older age and pseudophakia remained significant in the multivariable model. One explanation for this finding is the possibility of many uncoded pseudophakes in the dataset. It is also possible that providers may forgo gonioscopy during initial glaucoma evaluation for pseudophakic patients, as lens extraction dramatically reduces the risk of PACG.⁶ Gonioscopy may also be performed in patients with natural lenses to confirm presence of angle closure, which could provide an indication for earlier lens extraction. Lastly, gonioscopy requires patient cooperation and longer time at the slit lamp. Older patients are more likely to have comorbidities that may make them less tolerant of gonioscopy or deter providers from attempting it.^35,36

Our study found that patients residing outside of the Northeast region were less likely to receive gonioscopy. While it remains unclear whether this difference is related to practice or billing patterns, our results are consistent with prior studies that reported insured patients in the Northeast region are more likely to be detected with ANA prior to developing PACG..^37,38 There is also a higher rate of selective laser trabeculoplasty in the Northeast than in other regions, which would require the confirmation of an open angle.³⁹ Stricter adherence to recommended guidelines and higher diagnoses of glaucoma in the Northeast may be due to a high concentration of academic medical centers and fellowship trained glaucoma specialists.^40–43 Two studies comparing academic and non-academic sites found an 85% gonioscopy rate at a single academic clinic compared to a 53% gonioscopy rate among various private clinics.^13,14

Our study had several limitations. First, we analyzed healthcare claims data; therefore, procedures and diagnoses that were not coded or were miscoded were not observable in our dataset. This may bias our study sample by artificially lowering our gonioscopy estimates. Miscoding and/or undercoding of lens status may have also led to an underestimation of the effect of lens status. Second, we did not have access to granular clinical data on motivating factors for gonioscopy. However, an alternative database with biometric data on a very large, multiracial cohort is currently unavailable. Third, analyses were conducted on a patient-level instead of an eye-level due to lack of laterality data in ICD9, although gonioscopy is typically performed bilaterally. Finally, patients with unknown race also comprised a not insignificant proportion of our final cohort (3.9%), which may have affected racial associations with gonioscopy.

In conclusion, gonioscopy appears to be underperformed at initial glaucoma evaluation. Current practice patterns may contribute to racial disparities in angle closure outcomes, especially among racial groups in whom the burden of PACG is less studied, such as Blacks and Hispanics. While gonioscopy remains the clinical standard that should be performed in all glaucoma evaluations, our study supports the need for more convenient clinical methods to evaluate the angle. One example is anterior segment OCT (AS-OCT) imaging, which is less expertise dependent but can still detect eyes with gonioscopic angle closure and identify patients at higher risk for PACG. Therefore, emerging tools and methods could help improve provider adherence to standards of care and aid in earlier and more accurate disease detection and treatment.