Relationships among Clinical Factors and Patient-reported Outcome Measures in Adults with Convergence Insufficiency

Abstract

Significance:

When exploring relationships among clinical measures and patient-reported outcome measures in adults with convergence insufficiency, worse symptoms (Convergence Insufficiency Symptom Survey score) appeared to be correlated with worse Reading Function domain score (Adult Strabismus-20 quality-of-life questionnaire). After treatment, improved symptoms were associated with improved reading function quality of life.

Purpose.

To explore relationships between clinical measures and patient-reported outcome measures in adults undergoing treatment for symptomatic convergence insufficiency.

Methods.

In a prospective multicenter observational study we evaluated adults with symptomatic convergence insufficiency (i.e., clinical measures of near exodeviation, receded near point of convergence, reduced near positive fusional vergence; Convergence Insufficiency Symptom Survey (CISS) score ≥ 21). Fifty-seven participants treated with vision therapy/exercises (n=35) or base-in prism (n=22) were analyzed. Spearman correlation coefficients (R) were used to assess associations among the 3 clinical measures and patient-reported outcome measures (CISS, Diplopia Questionnaire, four Adult Strabismus-20 quality-of-life domains) before treatment (baseline) and after 10 weeks and 1 year. Associations were interpreted to be present when the lower limit of the 95% confidence interval (CI) was moderate-to-strong [R ≥ 0.4].

Results.

Among multiple exploratory analyses, the only moderate-to-strong baseline correlation was between worse CISS and worse Adult Strabismus-20 Reading Function scores (R=0.62, 95% CI: 0.43 to 0.76). Regarding change in measures with treatment, the only moderate-to-strong correlations were between improved CISS and improved AS-20 Reading Function scores for prism at 10 weeks (R=0.78, 95% CI: 0.52 to 0.91) and 1 year (R=0.85; (95% CI: 0.65 to 0.94), and for vision therapy/exercises at 1 year (R=0.78, 95% CI: 0.57 to 0.89).

Conclusions.

In exploratory analyses, we found positive correlations between CISS symptom scores and reading function quality-of-life scores. The absence of correlations between symptoms and individual clinical measures is consistent with clinical experience, that in convergence insufficiency, symptoms and clinical findings can be discordant.

Convergence insufficiency is a common binocular vision disorder typically characterized by a larger exodeviation at near (with or without manifest exotropia) than at distance, a receded near point of convergence, and decreased positive fusional vergence at near.^1,2 The reported prevalence of convergence insufficiency varies but has been estimated to be 5.5% in a population with a mean age of 30.5 years³ increasing to 21.5% in those 60 years and older.⁴ Adult-onset convergence insufficiency is reported to be the second most common type of newly diagnosed strabismus in adults.⁵

Convergence insufficiency is associated with a variety of symptoms including near blur, diplopia, headaches, eyestrain, and difficulty with reading or performing other tasks up close.^6,7 Reading challenges include reports of words moving on the page, loss of place, reading slowly, and loss of concentration.^6,7 Although convergence insufficiency has potential to adversely impact quality of life, the correlation between symptoms and quality of life has not been studied.

The aim of this present study was to explore associations among clinical measures, symptoms, and quality of life both before and after treatment in adults with symptomatic convergence insufficiency.

METHODS

The Pediatric Eye Disease Investigator Group (PEDIG) conducted this prospective multicenter observational study to describe clinical characteristics, treatments, and one-year outcomes of adults with symptomatic convergence insufficiency. Sample size was not set a priori, but was determined by the number of participants enrolled over a 2-year period. Choice of treatment was at investigator discretion. In addition to clinical measures, standardized symptom and quality-of-life measures were collected. Because the overall aim of this observational study was to provide pilot data on the numbers and characteristics of adults with convergence insufficiency seen by PEDIG investigators, with some preliminary data on treatment preferences and outcomes, there are important differences from a randomized trial including: treatments were not standardized, there was no attempt to measure adherence to treatment, and specifics of vision therapy/exercise protocols and procedures used were not collected.

This study was supported by the National Eye Institute and conducted in accordance with the Declaration of Helsinki at 19 clinical sites (listed on ClinicalTrials.gov⁸). The protocol is available on the PEDIG website.⁹ The protocol and Health Insurance Portability and Accountability Act-compliant informed consent forms were approved by the associated institutional review boards. All participants provided written informed consent.

Participants

Eligible participants were 18 years or older with a diagnosis of symptomatic convergence insufficiency and about to start a new treatment with either vision therapy/exercises, prism, extraocular muscle surgery, or botulinum toxin. Few participants were treated with surgery or botulinum toxin; thus, only those prescribed vision therapy/exercises or prism were included in this analysis. Primary inclusion criteria were near exodeviation ≥4Δ that was at least 4Δ larger than at distance (by prism and alternate cover test), distance exodeviation ≤15Δ, receded (≥ 6 centimeters) near point of convergence, insufficient positive fusional vergence at near (hereafter not specifying “at near”) (blur point [break, if no blur reported] <20Δ or failing Sheard’s criteria^{10, 11}), and a score of ≥ 21 points on the Convergence Insufficiency Symptom Survey (CISS⁶). Each measure was tested once at each examination. Table 1 lists additional eligibility and exclusion criteria.

Table 1.

Eligibility criteria.

The following criteria had to be met for the patient to be enrolled in the study:
1. Age ≥ 18 years (adult onset of convergence insufficiency not required)
2. No strabismus surgery within the past 10 years
3. Convergence Insufficiency Symptom Survey (CISS) score ≥21 points
4. Near exodeviation ≥4Δ and at least 4Δ larger than at distance by PACT
5. Distance exodeviation ≤15Δ by PACT
6. Vertical deviation ≤2Δ at distance and near by PACT
7. No constant exotropia at distance or near
8. Reduced positive fusional vergence at near (<20 or fails Sheard’s criterion that the positive fusional vergence measures less than twice the magnitude of the near phoria)
9. Near point of convergence ≥6 cm break
10. Visual acuity 20/50 or better in both eyes by ETDRS or Snellen
11. No paralytic strabismus (e.g., III, IV, or VI cranial nerve palsies, skew deviation, Duane syndrome)
12. No restrictive strabismus (e.g., blowout fracture, thyroid eye disease, post scleral buckle, Brown syndrome)
13. No monocular diplopia
14. No paretic strabismus, thyroid eye disease, myasthenia gravis, chronic progressive external ophthalmoplegia, or eye movement abnormalities associated with known neurological disease. Patients with Parkinson’s disease can be enrolled if their deviation is non-paretic.
15. No inferior or superior oblique overaction of 2+ or greater
16. If diplopic, has ability to fuse with prism in free space
17. Ability to understand and complete a survey
18. Investigator is initiating new treatment with prism, vision therapy/exercises, botulinum toxin injection, or surgery
19. If initiating treatment with botulinum toxin or surgery, planned injection or surgery to be within 60 days of enrollment
20. Single treatment modality planned (e.g., no combined prism and vision therapy/ exercises)
21. Treatment to be initiated has not been used within the prior year

Patient-reported Outcome Measures

Patient-reported outcome measures were used to assess symptoms and health-related quality of life. The severity of convergence insufficiency symptoms was quantified using the CISS,⁶ (Appendix 1) a 15-item validated questionnaire that evaluates frequency of symptoms experienced when reading or working at near using a 5-point scale (0 [never], 1 [infrequently], 2 [sometimes], 3 [fairly often], or 4 [always]). Item scores are summed to calculate the CISS score (range 0 [best] to 60 [worst]).⁶

Participants also completed the Diplopia Questionnaire (DQ),^{12, 13} (Appendix 2) a 7-item, self-administered questionnaire that quantifies diplopia severity. Two scores were calculated: overall (range: 0 [not diplopic] to 100 [always diplopic in all gaze positions]) using all 7 items and reading gaze only (range: 0 [not diplopic] to 40 [always diplopic for reading]).

To assess health-related quality of life, participants completed the Adult Strabismus-20 Questionnaire (AS-20),^14–17 (Appendix 3) a self-administered questionnaire with four unidimensional domains (General Function, Reading Function, Interactions, and Self-perception), and a 5-point response scale (never, rarely, sometimes, often, always). Each domain was Rasch-scored and rescaled from 0 (worst) to 100 (best) using published look-up tables.⁹

Clinical Measures

Three clinical measures were assessed through the current near optical correction: near exodeviation magnitude, near point of convergence, and positive fusional vergence. Exodeviation magnitude was determined by prism and alternate cover testing. Near point of convergence break was assessed using an Accommodative Convergence Rule (Gulden #1510 or similar, Gulden Ophthalmics, Elkins Park, PA) slowly moving the target (20/50-sized column of letters) until sustained diplopia was reported (or one eye drifted outward). Positive fusional vergence was measured using a horizontal prism bar (Gulden B-15) and a 20/50-sized column of letters at 40 centimeters. Prism magnitude was slowly increased until blur or diplopia was reported. The blur point was used for analysis when observed; otherwise, the break point was used.

Treatment Regimens

Of the 63 enrolled participants, 3 underwent strabismus surgery, 2 received botulinum toxin injections, and 1 changed treatment during the study; all 6 were excluded from analysis. The results reported herein are for the 57 participants prescribed vision therapy/exercises (n=35) or prism (n=22). Treatment was not standardized; the prism magnitude prescribed, and the specifics of the vision therapy prescribed such as dosage and procedures were at investigator discretion.

Follow-up Testing Procedures and Data Collection

Follow-up was 10 weeks (±3 weeks) and 1 year (±2 months) after the new treatment was prescribed. At each visit, the CISS, AS-20, and DQ were administered prior to clinical testing. Participants were asked to respond based on wearing their current refractive correction (including prism for those in the prism group). Near exodeviation magnitude, near point of convergence break, and positive fusional vergence were also measured.

Statistical Analysis

Our aim was to explore associations among clinical measures, symptoms, and quality of life both before and after treatment. Prior to analysis, the measures where a lower score indicates a better outcome (i.e., CISS, near point of convergence, DQ) were transformed, such that a better score, or in the case of change, improvement in score on both measures, would result in a positive correlation. Spearman rank correlation coefficients (R) and 95% confidence intervals (CI) were calculated to evaluate pairwise associations among clinical measures, quality of life, and symptoms prior to treatment. To be conservative in the interpretation of the results, an association was interpreted to be present only when the lower limit of the 95% CI indicated a moderate to strong association [R ≥ 0.4],¹⁸ rather than the point estimate alone. This very conservative interpretation of correlation coefficients was used because adjustment for multiplicity was not performed.

Associations evaluated at baseline were the clinical measures of near exodeviation magnitude, near point of convergence, and positive fusional vergence with 1) AS-20 General Function score; 2) AS-20 Reading Function score; 3) AS-20 Interactions score; 4) AS-20 Self-perception score; 5) CISS score; and 6) DQ score (both overall [0 to 100] and reading question only [0 to 40]). In addition, associations were evaluated between the CISS score and the four AS-20 domain scores.

Change in the clinical and patient-reported outcome measures was evaluated from enrollment to 10 weeks and from enrollment to 1 year. Associations evaluated at follow-up were change in near exodeviation magnitude, near point of convergence, and positive fusional vergence with change in: 1) AS-20 General Function score; 2) AS-20 Reading Function score; 3) AS-20 Interactions score; 4) AS-20 Self-perception score; 5) CISS score; and 6) DQ score (both overall and reading). We also evaluated whether change in the CISS score was associated with change in scores on each of the four AS-20 domains. The proportion of participants meeting the “symptom success criterion” (defined a priori as a CISS score <21 points that had improved by 9 or more points) was calculated for each treatment group at 10 weeks and 1 year.

Analyses of baseline associations were performed across all participants and analyses of change were performed separately for the two heterogeneous treatment groups (i.e., prism and vision therapy/exercises). Because our primary aim was to evaluate overall associations and not the effectiveness of specific treatment regimens (which are best addressed by rigorous randomized clinical trials), treatment subtypes defined by magnitude of prism or type of vision therapy/exercises prescribed are not reported separately.

RESULTS

Treatment Prescribed and Baseline Characteristics

Demographics, medical history, and baseline clinical characteristics are provided in Tables 2 and 3. For the 35 participants prescribed vergence therapy/exercises, 17 (49%) were prescribed therapy to be performed solely at home, 13 (37%) in-office therapy supplemented with home procedures, and 5 (14%) in-office therapy only. For the 22 participants prescribed base-in prism, the median magnitude of prism prescribed was 4Δ (range 1Δ to 8Δ).

Table 2.

Baseline Characteristics and medical history according to treatment prescribed at enrollment.

	Vision Therapy/Exercises n=35		Prism n=22
	n	%	n	%
Female	28	80%	16	73%
Race/ethnicity
Asian	5	14%	1	5%
Black/African American	3	9%	9	41%
Hispanic/Latino	5	14%	3	14%
White	21	60%	8	36%
More than one race	0	0%	0	0%
Unknown/not reported	1	3%	1	4%
Age (years)
18 to 24	11	31%	3	14%
25 to 34	11	31%	1	5%
35 to 44	5	14%	7	32%
45 to 54	3	9%	2	9%
55 to 64	3	9%	4	18%
65 to 74	2	6%	1	5%
75 to 84	0	0%	3	14%
85 to 91	0	0%	1	5%
Mean (SD), years	35.6 (15.1)		50.4 (20.2)
Median (IQR), years	30.3 (24.1 – 42.0)		45.9 (38.1 – 64.9)
Refractive correction worn at enrollment visit	23	66%	21	95%
Prior treatment for strabismus
None	27	77%	19	86%
Prism	4	11%	0	0%
Vision therapy/exercises	4	11%	2	9%
Coexisting neurological conditions
Parkinson’s disease	0	0%	1	5%
Stroke	2	6%	0	0%
Other conditions
Epiretinal membrane	0	0%	1	5%
Heart	1	3%	3	14%
Diabetes	2	6%	1	5%
Autoimmune disease	0	0%	1	5%
Significant head trauma	1	3%	3	14%
Othera	6	17%	1	5%
Tropia at near	3	9%	5	23%
Exodeviation at near by PACT (Δ)
Mean (SD)	12 (6)		14 (3)
Median (range)	10 (5 – 30)		14 (5 – 20)
Positive fusional vergence blur point (Δ) b
Mean (SD)	11 (7)		12 (7)
Median (range)	10 (2 – 35)		12 (1 – 30)
Near point of convergence (cm) c
Mean (SD)	14 (6)		15 (6)
Median (range)	12 (6 – 32)		13 (7 – 27)

^aOther conditions in the vision therapy/exercise group include a single case of migraines. In the prism group, other medical conditions were: asthma (1), concussion (1), history of resected meningioma (1), spinal stenosis and hypothyroidism (1), hypothyroidism (1), and Chiari malformation with post-surgical traumatic brain injury (1).

^bThe break point was used if no blur was reported;

^c6 cm or worse required for eligibility; PACT = prism and alternate cover test; IQR = interquartile range; SD = standard deviation

Table 3.

Patient-reported outcome measures of symptoms and quality of life at enrollment according to treatment prescribed.

Scores	Vision Therapy/Exercises n=35	Prism n=22
Convergence Insufficiency Symptom Survey
Mean (SD)	38 (10)	37 (8)
Median (range) a	37 (21 – 60)	36 (23 – 54)
Diplopia Questionnaire (Overall)
Mean (SD)	30 (24)	28 (30)
Median (range) b	26 (0 – 100)	22 (0 – 100)
Diplopia Questionnaire (Reading)
Mean (SD)	16 (14)	15 (15)
Median (range) c	20 (0–40)	20 (0–40)
Adult Strabismus-20 General Function
Mean (SD)	56 (24)	56 (25)
Median (range) b	60 (17 – 92)	56 (0 – 92)
Adult Strabismus-20 Reading Function
Mean (SD)	42 (22)	47 (23)
Median (range) b	41 (7 – 90)	45 (13 – 89)
Adult Strabismus-20 Interactions
Mean (SD)	92 (17)	92 (10)
Median (Range) b	100 (10 – 100)	97 (64 – 100)
Adult Strabismus-20 Self-perception
Mean (SD)	83 (25)	85 (18)
Median (range) b	91 (0 – 100)	95 (48 – 100)

SD = standard deviation;

^aPossible range 0 – 60 points;

^b-Possible range 0 – 100 points;

^c-Possible range 0 – 40 points

Associations at Enrollment

At enrollment prior to starting treatment, there was a moderate to strong association between worse (higher) CISS scores and worse (lower) AS-20 Reading Function scores (R = 0.62, 95% CI: 0.43 to 0.76) (Table 4, Figure 1), with the association between the CISS and AS-20 General Function being the next strongest (R = 0.50, 95% CI: 0.28 to 0.68; Table 4) but not meeting our strict definition of moderate to strong correlation. There were no other moderate to strong associations found between clinical measures (near exodeviation magnitude, near point of convergence, and positive fusional vergence) and patient-reported outcome measures (AS-20 domain scores, CISS score, and DQ score) among the multiple associations explored at enrollment (Table 4).

Table 4.

Correlations between measures at enrollment for participants subsequently treated with either vision therapy/exercises (n=35) or prism (n=22).

Measure 1	Measure 2	Correlation (95% CI)a
Near Point of Convergence (cm)	AS-20 General Function domain score	0.05 (−0.21 to 0.31)
	AS-20 Reading Function domain score	−0.07 (−0.33 to 0.19)
	AS-20 Interactions domain score	0.25 (−0.01 to 0.48)
	AS-20 Self-perception domain score	0.22 (−0.04 to 0.46)
	DQ Overall score	0.28 (0.02 to 0.50)
	DQ Reading score	0.15 (−0.11 to 0.40)
	Positive Fusional Vergence	0.35 (0.10 to 0.56)
	CISS score	−0.03 (−0.29 to 0.23)
Positive Fusional Vergence at near (Δ)	AS-20 General Function domain score	−0.03 (−0.29 to 0.23)
	AS-20 Reading Function domain score	−0.05 (−0.31 to 0.21)
	AS-20 Interactions domain score	0.05 (−0.21 to 0.31)
	AS-20 Self-perception domain score	0.12 (−0.15 to 0.37)
	DQ Overall score	0.15 (−0.11 to 0.40)
	DQ Reading score	0.24 (−0.02 to 0.47)
	CISS score	0.00 (−0.26 to 0.26)
CISS score (points)	AS-20 General Function domain score	0.50 (0.28 to 0.68)
	AS-20 Reading Function domain score	0.62 (0.43 to 0.76)
	AS-20 Interactions domain score	0.10 (−0.17 to 0.35)
	AS-20 Self-perception domain score	0.10 (−0.16 to 0.36)
	DQ Overall score	0.17 (−0.09 to 0.41)
	DQ Reading score	0.22 (−0.04 to 0.46)
Near Exodeviation Magnitude (Δ)	AS-20 General Function domain score	0.20 (−0.06 to 0.44)
	AS-20 Reading Function domain score	0.23 (−0.03 to 0.47)
	AS-20 Interactions domain score	0.17 (−0.09 to 0.41)
	AS-20 Self-perception domain score	0.21 (−0.06 to 0.44)
	DQ Overall score	0.13 (−0.13 to 0.38)
	DQ Reading score	0.21 (−0.05 to 0.45)
	CISS score	0.21 (−0.06 to 0.44)

AS-20 = Adult Strabismus Survey, CISS = Convergence Insufficiency Symptom Survey, DQ = Diplopia Questionnaire, NPC = Near point of convergence break point, PACT = prism and alternate cover test; CI = confidence interval; Δ = prism diopters.

Bold indicates a moderate or strong correlation. Interpretation of the magnitude of R was based on the entire CI rather than the point estimate alone. To be conservative, an association was interpreted as being present when both the upper and lower limits of the 95% CI indicated a moderate to strong association (R ≥ 0.4).¹⁸

^aA positive correlation indicates a relationship of worse Measure 1 with worse Measure 2

^bblur point (or break, if no blur reported).

Figure 1.

Association of Convergence Insufficiency Symptom (CISS) survey (CISS) scores with AS-20 Reading Function scores at enrollment. (CISS scores have been transformed so that a positive correlation indicates better CISS scores associated with better AS-20 scores.)

Change and Associations at Follow-up: Treatment with Vision Therapy/Exercises

Of the 35 participants prescribed vision therapy/exercises, follow-up data were available for 34 (97%) at 10 weeks and 27 (77%) at 1 year. Change in clinical measures and patient-reported outcome measures, with 95% CIs, are shown in Table 5. On average, there were improvements in both clinical and patient-reported outcome measures by 10 weeks with additional improvements by 1 year, with the exception of AS-20 Interactions at 1 year (Table 5). At 10 weeks, 9% (3 of 34; 95% CI: 2%–24%) met the symptomatic success criterion; 52% (14 of 27; 95% CI: 32%–71%) met the criterion by 1 year.

Table 5.

Mean improvement in clinical measures and patient-reported outcome measures from enrollment to 10 weeks and from enrollment to 1 year.

Clinical or Patient-Reported Outcome Measure	10 Weeks	1 Year
	Mean Improvement (95% CI)a	Mean Improvement (95% CI)a
Vision therapy/exercises prescribed at enrollment	N=34	N=27
AS-20 General Function (points)	10.4 (4.1 to 16.6)	22.5 (13.5 to 31.5)
AS-20 Reading Function (points)	11.9 (4.9 to 19.0)	27.3 (16.4 to 38.3)
AS-20 Interactions (points)	−0.1 (−2.8 to 2.5)	−0.3 (−3.1 to 2.6)
AS-20 Self-perception (points)	3.1 (−2.1 to 8.3)	2.6 (−3.7 to 9.0)
DQ Overall (points)	4.0 (−2.3 to 10.3)	14.9 (5.3 to 24.5)
DQ Reading (points)	2.3 (−1.6 to 6.2)	8.5 (2.0 to 15.1)
CISS (points)	7.9 (4.9 to 10.8)	15.7 (10.9 to 20.5)
NPC (cm)	4.4 (0.9 to 7.8)	6.9 (4.4 to 9.4)
Positive fusional vergence (Δ)b	5.7 (2.5 to 8.8)	10.1 (6.0 to 14.3)
Near exodeviation magnitude (Δ)	1.8 (0.9 to 2.8)	2.2 (0.8 to 3.6)
Prism prescribed at enrollment	N=21	N=19
AS-20 General Function (points)	25.9 (12.8 to 39.0)	29.5 (14.8 to 44.3)
AS-20 Reading Function (points)	25.4 (11.9 to 38.9)	33.4 (19.9 to 46.9)
AS-20 Interactions (points)	4.1 (0.8 to 7.4)	5.9 (2.0 to 9.9)
AS-20 Self-perception (points)	4.9 (−1.7 to 11.5)	7.8 (0.2 to 15.4)
DQ Overall (points)	14.1 (6.4 to 21.7)	18.4 (7.6 to 29.1)
DQ Reading (points)	7.6 (3.4 to 11.8)	8.4 (3.8 to 13.0)
CISS (points)	14.0 (7.2 to 20.7)	17.9 (10.1 to 25.7)
NPC (cm)	2.5 (−0.1 to 5.1)	0.6 (−3.6 to 4.9)
Positive fusional vergence (Δ)b	4.6 (−0.3 to 9.5)	3.2 (−1.1 to 7.5)
Near exodeviation magnitude (Δ)	2.3 (0.2 to 4.4)	1.9 (−0.9 to 8.7)

AS-20 = Adult Strabismus-20 Questionnaire, CISS = Convergence Insufficiency Symptom Survey, DQ = Diplopia Questionnaire, NPC = near point of convergence break point, Δ = prism diopters; CI = confidence interval

^aWhile not a comparative effectiveness study, clinical values are reported to provide the data used for analyses of associations. Positive values indicate an improvement in the score or the condition;

^bblur point (or break if no blur was reported)

Regarding associations between measures, no moderate to strong associations were found between change in the three clinical measures and change in patient-reported outcome measures over 10 weeks or over 1 year. Among the patient-reported outcome measures, improvement in the CISS score was moderately to strongly associated with an improved AS-20 Reading Function score at 1 year (R = 0.78, 95% CI: 0.57 to 0.89) (Table 6, Figure 2A). The moderate correlation between the vision therapy/exercise group’s performance on the CISS and the AS-20 General Function scores at 10 weeks (R = 0.59; 95% CI: 0.31 to 0.77) and 1 year (R = 0.59; 95% CI: 0.27 to 0.79) were the next strongest associations but did not meet our strict criterion for a moderate to strong association. This also was true for the positive association (R = 0.51; 95% CI: 0.21 to 0.73) between the CISS score and the AS-20 Reading Function score at 10 weeks. Other weaker associations are reported in Table 6.

Table 6.

Correlations between changes in measures 10 weeks and 1 year after enrollment for participants prescribed vision therapy/exercises.

Measure 1	Measure 2	10 Weeks (N=34)	1 Year (N=27)
		R (95% CI)a	R (95% CI)a
Near Point of Convergence (cm)	Change in AS-20 General Function score	0.00 (−0.34 to 0.34)	−0.11 (−0.47 to 0.28)
	Change in AS-20 Reading Function score	−0.09 (−0.42 to 0.26)	−0.34 (−0.64 to 0.05)
	Change in AS-20 Interactions score	−0.30 (−0.58 to 0.04)	−0.16 (−0.51 to 0.23)
	Change in AS-20 Self-perception score	−0.08 (−0.41 to 0.27)	0.03 (−0.35 to 0.41)
	Change in DQ Overall score	0.16 (−0.19 to 0.47)	−0.15 (−0.50 to 0.25)
	Change in DQ Reading score	0.23 (−0.12 to 0.52)	−0.09 (−0.45 to 0.30)
	Change in CISS score	0.02 (−0.32 to 0.36)	−0.27 (−0.59 to 0.12)
Change in Positive Fusional Vergence at near (Δ)b	Change in AS-20 General Function score	0.24 (−0.11 to 0.53)	0.11 (−0.28 to 0.47)
	Change in AS-20 Reading Function score	0.14 (−0.21 to 0.45)	−0.01 (−0.39 to 0.37)
	Change in AS-20 Interactions score	−0.06 (−0.39 to 0.28)	0.06 (−0.33 to 0.43)
	Change in AS-20 Self-perception score	0.13 (−0.22 to 0.45)	0.27 (−0.12 to 0.59)
	Change in DQ Overall score	0.26 (−0.08 to 0.55)	−0.12 (−0.48 to 0.27)
	Change in DQ Reading score	0.39 (0.05 to 0.64)	−0.03 (−0.40 to 0.36)
	Change in CISS score	0.23 (−0.12 to 0.52)	0.08 (−0.31 to 0.44)
Change in Near Exodeviation Magnitude (Δ)	Change in AS-20 General Function score	−0.31 (−0.59 to 0.03)	−0.24 (−0.57 to 0.15)
	Change in AS-20 Reading Function score	0.04 (−0.30 to 0.37)	−0.18 (−0.53 to 0.21)
	Change in AS-20 Interactions score	0.03 (−0.31 to 0.37)	−0.20 (−0.54 to 0.19)
	Change in AS-20 Self-perception score	−0.04 (−0.37 to 0.30)	0.11 (−0.28 to 0.47)
	Change in DQ Overall score	0.03 (−0.31 to 0.36)	0.24 (−0.16 to 0.57)
	Change in DQ Reading score	0.21 (−0.13 to 0.51)	0.28 (−0.12 to 0.59)
	Change in CISS score	−0.08 (−0.41 to 0.26)	−0.08 (−0.45 to 0.31)
Change in CISS score (points)	Change in AS-20 General Function score	0.59 (0.31 to 0.77)	0.59 (0.27 to 0.79)
	Change in AS-20 Reading Function score	0.51 (0.21 to 0.73)	0.78 (0.57 to 0.89)
	Change in AS-20 Interactions score	−0.22 (−0.52 to 0.13)	0.54 (0.20 to 0.76)
	Change in AS-20 Self-perception score	0.06 (−0.28 to 0.39)	0.17 (−0.22 to 0.52)
	Change in DQ Overall score	0.21 (−0.14 to 0.51)	0.39 (0.01 to 0.67)
	Change in DQ Reading score	0.24 (−0.11 to 0.53)	0.43 (0.07 to 0.70)

AS-20 = Adult Strabismus-20 Questionnaire, CISS = Convergence Insufficiency Symptom survey, DQ = Diplopia Questionnaire, CI = confidence interval; Δ = prism diopters

^aPositive correlation values indicate an improvement in both scores or both clinical measures;

^bblur point (or break, if no blur reported)

Bold indicates a positive, moderate or strong correlation (i.e., improvement in both factors). Interpretation of the magnitude of R was based on the entire CI rather than the point estimate alone. To be conservative, an association was interpreted as being present when both the upper and lower limits of the 95% CI indicated a moderate to strong association (R ≥ 0.4).¹⁸

Figure 2.

Association of Convergence Insufficiency Symptom (CISS) survey (CISS) scores with AS-20 Reading Function scores at 10 weeks and 1 year. (A) after 1 year of vision therapy/exercises, (B) after 10 weeks of prism treatment, and (C) after 1 year of prism treatment. (CISS scores have been transformed so that a positive correlation indicates better CISS scores are associated with better AS-20 scores.)

Change and Associations Found at Follow-up: Treatment with Prism

Ten-week and 1-year follow-up data were available for 21 (95%) and 19 (86%) of 22 participants prescribed prism, respectively. Table 5 shows changes in clinical and patient-reported outcome measures. Overall, there were improvements by 10 weeks that were sustained at 1 year (Table 5). Ten of the 21 (48%; 95% CI: 26%–70%) participants met the symptomatic success criterion by 10 weeks, and 10 of 19 (53%; 95% CI: 29%–76%) by 1 year.

There were no moderate to strong associations found between change in any of the three clinical measures and change in patient-reported outcome measures. However, among the patient-reported outcomes, an improved CISS score was moderately to strongly associated with an improved AS-20 Reading Function score at 10 weeks (R = 0.78; 95% CI: 0.52 to 0.91) and 1 year (R = 0.85; 95% CI: 0.65 to 0.94) (Table 7, Figures 2B–2C). The moderate association between the CISS and AS-20 General Function scores at 10 weeks (R = 0.69; 95% CI: 0.37 to 0.87) and 1 year (R = 0.71; 95% CI: 0.39 to 0.88) did not meet our conservative criterion that both ends of the confidence intervals be >0.40. Other (weaker) associations are reported in Table 7.

Table 7:

Correlations between changes in measures 10 weeks and 1 year after enrollment for participants prescribed base-in prism.

Measure 1	Measure 2	10 Weeks (N=21)	1 Year (N=19)
		R (95% CI)a	R (95% CI)a
Change in Near Point of Convergence (cm)	Change in AS-20 General Function score	−0.01 (−0.45 to 0.43)	0.15 (−0.34 to 0.57)
	Change in AS-20 Reading Function score	0.13 (−0.33 to 0.54)	0.46 (−0.00 to 0.76)
	Change in AS-20 Interactions score	0.30 (−0.16 to 0.66)	0.39 (−0.09 to 0.73)
	Change in AS-20 Self-perception score	0.38 (−0.07 to 0.70)	0.44 (−0.03 to 0.75)
	Change in DQ Overall score	0.21 (−0.26 to 0.59)	0.19 (−0.31 to 0.60)
	Change in DQ Reading score	−0.13 (−0.54 to 0.33)	0.03 (−0.44 to 0.49)
	Change in CISS score	0.18 (−0.28 to 0.58)	0.31(−0.18 to 0.68)
Change in Positive Fusional Vergence at near (Δ)	Change in AS-20 General Function score	−0.27 (−0.63 to 0.18)	0.18 (−0.30 to 0.59)
	Change in AS-20 Reading Function score	0.10 (−0.35 to 0.51)	0.36 (−0.11 to 0.70)
	Change in AS-20 Interactions score	−0.01 (−0.44 to 0.42)	0.08 (−0.39 to 0.52)
	Change in AS-20 Self-perception score	0.02 (−0.41 to 0.45)	0.13 (−0.35 to 0.55)
	Change in DQ Overall score	−0.30 (−0.65 to 0.15)	−0.12 (−0.55 to 0.35)
	Change in DQ Reading score	−0.02 (−0.45 to 0.41)	−0.15 (−0.57 to 0.32)
	Change in CISS score	−0.04 (−0.46 to 0.40)	0.38 (−0.09 to 0.71)
Change in Near Exodeviation Magnitude (Δ)	Change in AS-20 General Function score	0.49 (0.07 to 0.76)	0.27 (−0.21 to 0.64)
	Change in AS-20 Reading Function score	0.54 (0.14 to 0.79)	0.14 (−0.34 to 0.56)
	Change in AS-20 Interactions score	0.03 (−0.41 to 0.46)	−0.11 (−0.54 to 0.36)
	Change in AS-20 Self-perception score	0.15 (−0.30 to 0.55)	−0.29 (−0.66 to 0.19)
	Change in DQ Overall score	−0.06 (−0.48 to 0.38)	−0.09 (−0.52 to 0.38)
	Change in DQ Reading score	0.04 (−0.39 to 0.47)	−0.02 (−0.47 to 0.44)
	Change in CISS score	0.56 (0.18 to 0.80)	0.48 (0.04 to 0.77)
Change in CISS score (points)	Change in AS-20 General Function score	0.69 (0.37 to 0.87)	0.71 (0.39 to 0.88)
	Change in AS-20 Reading Function score	0.78 (0.52 to 0.91)	0.85 (0.65 to 0.94)
	Change in AS-20 Interactions score	0.23 (−0.22 to 0.60)	−0.20 (−0.60 to 0.28)
	Change in AS-20 Self-perception score	0.43 (0.00 to 0.73)	0.04 (−0.42 to 0.49)
	Change in DQ Overall score	0.28 (−0.17 to 0.64)	0.29 (−0.18 to 0.66)
	Change in DQ Reading score	0.21 (−0.24 to 0.59)	0.14 (−0.33 to 0.56)

AS-20 = Adult Strabismus-20 Questionnaire, CISS = Convergence Insufficiency Symptom survey, DQ = Diplopia Questionnaire, NPC = near point of convergence break point, PFV = positive fusional vergence blur point (or break, if no blur reported); CI = confidence interval; Δ = prism diopters;

^aPositive correlation values indicate an improvement in both scores or both clinical measures

Bold indicates a positive, moderate or strong correlation (i.e., improvement in both factors). Interpretation of the magnitude of R was based on the entire CI rather than the point estimate alone. To be conservative, an association was interpreted as being present when both the upper and lower limits of the 95% CI indicated a moderate to strong association (R ≥ 0.4).¹⁸

DISCUSSION

In this prospective observational study of adults with symptomatic convergence insufficiency, we explored multiple potential associations among clinical measures and patient-reported outcome measures. Applying stringent criteria, the only moderate to strong association at enrollment was between worse symptoms (CISS score) and worse reading-related quality of life (AS-20 Reading Function domain). We also found that an improvement in symptoms (CISS score) was associated with an improvement in Reading Function quality of life (AS-20 domain score) when measured at 10 weeks and 1-year post-initiation of both vision therapy and prism treatments. Improvements in symptoms and quality of life with base-in prism appeared to be evident sooner, reflected in gains at 10 weeks (and sustained at 1 year), whereas improvements with vision therapy/exercises appeared to continue after 10 weeks and be more apparent at the 1-year follow-up visit.

While strabismus in general has been reported to impact quality of life,¹⁴ the impact of convergence insufficiency (with or without manifest exotropia) does not appear to have been previously studied. Indeed, none of the 12 randomized trials comprising a recent Cochrane systematic review¹ of non-surgical interventions for convergence insufficiency included a quality-of-life measure. The present study confirms that adult convergence insufficiency predominantly impacts reading function quality of life (assessed using the AS-20 questionnaire), unlike other, more visually obtrusive types of strabismus that may also impact social and interaction domains.

We found no relationships evident between any of the three patient-reported outcome measures (CISS score, Diplopia Questionnaire, AS-20 questionnaire) and the clinical measures of exodeviation magnitude, near point of convergence, or positive fusional vergence. Other investigators have similarly reported a lack of correlation between clinical measures and CISS score. Pang and colleagues¹⁹ reported that exodeviation magnitude, near point of convergence, and positive fusional vergence were not associated with the CISS score before or after base-in prism treatment in presbyopic adults with symptomatic convergence insufficiency. Bade and colleagues²⁰ also found no significant correlations between these three clinical measures and the CISS score in a randomized trial of 9- to 17-year-old children with convergence insufficiency. These findings of discordance between quantitative clinical measures of convergence insufficiency and symptoms are consistent with experience in clinical practice where some patients can be profoundly symptomatic despite little to no clinical abnormality, while other patients categorically deny symptoms despite clinical measures that are extremely abnormal. Convergence insufficiency is not the only medical condition where there is discordance between disease severity and symptom reports, with similar findings reported for ocular conditions such as dry eye^{21, 22} and various systemic diseases.^23–26 In addition, demographic, social, cultural, environmental, and psychological factors are known to influence symptom reports.²⁶ For convergence insufficiency, a disorder that is only present at near, we would also predict that the amount and type of near work performed are likely to play a role in symptom severity, and suppression, if present, could also mitigate symptoms. The relationship between clinical measures of convergence insufficiency and symptom severity is worthy of further study.

Perhaps unsurprisingly, given there were no correlations among clinical measures and symptoms (CISS score) at enrollment, we also found no moderate to strong relationships between changes in these measures after treatment. A similar finding was recently reported by Alvarez and coworkers²⁷ in a randomized trial of active versus placebo vision therapy in 50 adults with convergence insufficiency; they found no correlation between changes in clinical measures and changes in CISS score. In addition, a randomized trial of young adults with convergence insufficiency conducted by Scheiman and colleagues²⁸ found that improvements in near point of convergence and positive fusional vergence did not correspond to improvements in symptoms (CISS score). This lack of correlation between improvements in clinical measures and patient-reported outcome measures has also been reported in a randomized trial of children with convergence insufficiency.²⁹ However, because the goal of treatment for convergence insufficiency is to improve symptoms, the lack of correlation between clinical parameters and patient-reported measures should not be interpreted as evidence of a lack of value for treatment.

Recent randomized clinical trials of vision therapy for adults with convergence insufficiency^{27, 28} found a greater magnitude of change in clinical parameters than found in the present study (Table 5). This is likely attributable (at least in part) to the real-world but non-standardized treatment approaches used in this observational study versus the standardized, office-based treatment protocols used in the randomized trials.

There are limitations to our study. Sample size was relatively small which resulted in wide confidence intervals, thereby limiting our ability to identify associations meeting our criteria for moderate to strong relationships. In addition, our study cohort does not represent the spectrum of adults with symptomatic convergence insufficiency and therefore our results are only generalizable to those with similar clinical and demographic characteristics. For example, approximately two-thirds of the adult participants in the vision therapy/exercise group were younger than 35 years of age, with few in the age range with the highest prevalence of convergence insufficiency (older than 60 years).⁴ It is unknown whether the associations we found would differ (stronger or weaker) in older persons. In addition, because our study was not a randomized clinical trial with standardized treatment protocols, a control group, or measures of treatment adherence, we cannot comment on treatment effectiveness, or rule out a placebo component of the treatments. While prior clinical trials^1,27,28 have shown differences in outcomes for treatment sub-groups (e.g., home-based and office-based vision therapy/exercises), our modest sample size precluded such sub-group analyses and thus we do not know whether associations would differ among treatment sub-types. The small number of non-diplopic participants in our cohort limited our ability to evaluate the differential effect of presence of diplopia. Because we were conducting exploratory analyses, we did not adjust for multiplicity. However, we were very conservative in that we only considered moderate to strong associations to be those where both the upper and lower limits of the 95% CI did not include a weak correlation (i.e., both the upper and lower limits were consistent with a moderate to strong correlation). Thus, it is possible that other moderate to strong correlations were present but were not identified as such in the present study because the confidence intervals were wide. Finally, because it is theoretically possible that agreement between the CISS and AS-20 Reading Function domain scores may be due to measuring the same trait (both contain reading items), we conducted exploratory factor analyses (data not shown) that confirmed the two instruments each measure different traits (symptoms vs. quality of life). However, this exploratory factor analysis also suggested multidimensionality in the CISS which should be addressed in future studies of larger adult populations.

In conclusion, in exploratory analyses we found that worse symptom severity correlated with worse reading function quality-of-life in adults with convergence insufficiency, suggesting that patient-reported outcome measures may be useful in the assessment and management of adults with convergence insufficiency. The absence of correlations between symptoms and individual clinical measures is consistent with clinical experience that there can be discordance between symptoms and clinical findings in adults with convergence insufficiency. The present study findings are helpful for generating hypotheses for future research and for aiding interpretation of randomized trials^27–29 reporting discordance between symptoms and clinical measures.

Supplementary Material

PEDIG appendices final

Appendix 1, available at [LWW insert link]: Convergence Insufficiency Symptom Survey

Appendix 2, Diplopia Questionnaire, available at [LWW insert link].

Appendix 3, available at [LWW insert link]: The Adult Strabismus 20 (AS-20) Quality of Life Questionnaire Consisting of Four Distinct Domains: Self-Perception, Interactions, General Function, Reading Function.