To evaluate changes in health-related quality of life (HRQOL) in adult strabismus patients classified as surgical failures by standard motor and diplopia criteria.
Prospective cohort study evaluating outcomes.
Adults undergoing strabismus surgery in a single clinical practice, with preoperative and 1-year-postoperative Adult Strabismus-20 HRQOL questionnaires, were included. Motor and diplopia criteria were applied to classify outcomes (success, partial success, or failure). For those classified as failure, the medical record of the 1-year examination was reviewed to determine whether the patient reported subjective improvement. We evaluated improvement in HRQOL, defined as exceeding 95% limits of agreement on at least 1 of the 4 Adult Strabismus-20 domains. We compared proportions exceeding 95% limits of agreement in those reporting subjective improvement vs those who did not.
Forty of 227 patients (18%) were classified as failure by motor and diplopia criteria, with 39 of 40 able to exceed Adult Strabismus-20 95% limits of agreement. Overall, 21 of 39 (54%) showed improved HRQOL by exceeding 95% limits of agreement on at least 1 of the 4 Adult Strabismus-20 domains (54% vs predicted 10% by chance alone; P < .0001). Twenty-five patients (64%) reported subjective improvement, of whom 16 (64%) showed improved HRQOL exceeding 95% limits of agreement.
Many apparent surgical failures report subjective improvement, often reflected in improved HRQOL scores. We propose incorporating quantitative HRQOL criteria into the assessment of strabismus surgery outcomes, defining success as either meeting motor and diplopia criteria or showing improvement in HRQOL beyond test-retest variability.
In adults, strabismus surgery outcomes are typically assessed by evaluating the angle of deviation. In a previous study of outcomes we concluded that combining motor criteria and diplopia criteria provided a more representative assessment of surgical outcomes than using either criterion on its own. Combining motor and diplopia criteria raises the standard for success and reduces the likelihood that an unsatisfactory outcome would be misclassified as “success.” Nevertheless, combining criteria also means that patients experiencing significant improvement in motor alignment alone, and meeting criteria for motor success, could be classified as overall failures based on the presence of diplopia. In addition, many patients with profound ocular dysmotility preoperatively experience significant improvement in alignment and diplopia following surgery, yet without meeting criteria for success. The aim of this study was to assess patients classified as surgical failures by motor and diplopia criteria for evidence of subjective improvement following surgery and to analyze health-related quality of life (HRQOL) scores to determine whether any documented subjective improvement was reflected by changes in HRQOL.
Institutional Review Board approval for the review of participants’ clinical and questionnaire data in this retrospective cohort study evaluating outcomes was obtained prior to commencement of the study from the Institutional Review Board at Mayo Clinic, Rochester, Minnesota. All procedures and data collection were conducted in a manner compliant with the Health Insurance Portability and Accountability Act. All research procedures adhered to the tenets of the Declaration of Helsinki.
We retrospectively identified adult patients undergoing strabismus surgery in a single clinical practice who had completed the Adult Strabismus-20 questionnaire preoperatively (window: 1–28 days) and 1 year postoperatively (window: 5 months to 2 years). Postoperative questionnaire data were taken from the examination nearest to 1 year following surgery. Patients with any type of diplopic and nondiplopic strabismus were included; no exclusions were made based on diagnosis. Patients were not included if they were unable to read or understand English, had severe cognitive impairment, or had undergone additional extraocular muscle surgery before the 1-year outcome. All patients self-completed the Adult Strabismus-20 questionnaire, typically while in the waiting area and before any clinical testing was performed.
Adult Strabismus-20 Questionnaire
The Adult Strabismus-20 questionnaire is a strabismus-specific questionnaire developed to assess the effects of strabismus on HRQOL in adults. In Rasch analysis of the Adult Strabismus-20 4 distinct domains were identified: Self-Perception, Interactions, Reading Function, and General Function (full questionnaire freely available at www.pedig.net , accessed October 21, 2015). Each of the 4 domains is scored independently using Rasch scoring methods and converted to a 0-100 score (worst to best HRQOL) for easier interpretation (scoring lookup table freely available at www.pedig.net , accessed October 21, 2015). Adult Strabismus-20 scores were calculated for each patient, for each of the 4 domains, at both the preoperative and the 1-year-postoperative examination.
As part of the standard clinical examination, the angle of deviation was measured in prism diopters (PD) using the simultaneous prism and cover test and the prism and alternate cover test at distance (3 meters) and near (1/3 meter) fixation, in habitual refractive correction. If the patient had poor visual acuity that precluded accurate prism cover test measurements, Krimsky measurements were used. If prism correction was worn, and only in-prism measurements were available, the simultaneous prism and cover test in prism correction was added to the prism strength to represent the underlying manifest angle of deviation.
All patients completed a diplopia questionnaire as part of their clinical examination, rating the frequency of any diplopia in various gaze positions (straight-ahead distance, reading, upgaze, downgaze, right gaze, left gaze, any other position) as noticed over the past week. For each gaze position the frequency of any diplopia was rated as either never, rarely, sometimes, often, or always. This previously reported questionnaire was scored using a data-driven scoring algorithm on a 0 (no diplopia) to 100 (constant diplopia) scale (questionnaire and scoring algorithm freely available at www.pedig.net , accessed October 21, 2015).
Classification of Postoperative Outcomes
Postoperative outcomes were classified using previously described motor and diplopia criteria. For motor criteria, simultaneous prism and cover test angle data were used in order to capture the manifest misalignment. Postoperative outcomes were classified as “failure” if 1 of the following criteria was met: (1) simultaneous prism and cover test was 15 PD or more (horizontal or vertical) at distance or near; (2) diplopia or visual confusion was present more than “sometimes” straight ahead at distance or for reading (unless atypical diplopia due to decompensated childhood strabismus was present preoperatively, in which case diplopia was allowed postoperatively); (3) the patient was wearing an occlusive patch or Bangerter foil. Outcomes were classified as partial success if simultaneous prism and cover test was ≤15 PD (horizontal and vertical) at distance and near, and diplopia or visual confusion was present never, rarely, or sometimes. Correction of diplopia with prism was allowed for classification as partial success. Outcomes were classified as success if simultaneous prism and cover test was <10 PD (horizontal and vertical) at distance and near, and diplopia or visual confusion was present never or only rarely.
For patients whose 1-year-postoperative outcome was classified as failure, the history section of the medical record was reviewed to determine whether the patient had experienced any improvement following surgery—for example, statements by the patient that double vision or alignment was improved compared with preoperatively (for specific quotations see results). This assessment was made independently, without knowledge of preoperative or postoperative Adult Strabismus-20 HRQOL scores.
Change in preoperative to postoperative Adult Strabismus-20 scores was calculated and median preoperative to postoperative scores were compared using the signed rank test. Improvement in HRQOL for an individual subject was defined as change in score exceeding the 95% limits of agreement on at least 1 of the 4 Adult Strabismus-20 domains. The 95% limits of agreement define the limits within which 95% of differences in score due to test-retest variability should lie. Therefore a change in score exceeding these limits is likely to represent a meaningful change. In our present study, the 95% limits of agreement were calculated using 1.96 standard deviation of previously reported test-retest differences to define the limits. The 95% limits of agreement for the 4 Adult Strabismus-20 domains, calculated from these data, were 30.37 points for Self-Perception, 19.32 for Interactions, 24.75 for Reading Function, and 27.66 for General Function. Any patient unable to improve on at least 1 Adult Strabismus-20 domain (preoperative scores too high) was excluded from further analysis.
To determine whether reported subjective improvement in the medical record was reflected by a measured improvement in HRQOL, we compared the proportion of patients showing improved HRQOL in those who reported subjective improvement and those who did not, using Fisher exact tests.
Two hundred and twenty-seven adults undergoing surgery and with preoperative and 1-year-postoperative Adult Strabismus-20 and diplopia questionnaire data were identified. One hundred and forty-two (63%) were female and 96% reported their race as white. Median age at the 1-year examination was 53 (range 18-88) years. Overall, 40 of 227 (18%) were classified as failures, 128 (56%) were successes, and 59 (26%) were partial successes. Data from 4 of 40 patients (10%) have been reported previously in prior studies.
Reasons for Surgical Failure
Across 40 failures, 32 (80%) were diplopic preoperatively, 7 had no diplopia preoperatively, and 1 had atypical diplopia associated with loss of suppression. Preoperative strabismus types were cranial nerve palsies (n = 15), restrictive (n = 11), idiopathic/childhood onset (n = 10), sensory (n = 3), and other neuro (n = 1). For 34 of 40 failures (85%) the reason for surgical failure was diplopia, rated as “often” or “always” for straight-ahead distance or for reading; the remaining 6 patients (15%) failed for a simultaneous prism and cover test of 15 PD or more at distance or near. Subsequent to the 1-year follow-up examination, 19 of 40 failures (48%) went on to receive additional surgery (18 incisional surgery and 1 Botulinum toxin injection).
Median Improvement in Adult Strabismus-20 Scores for Failures
On average, preoperative and postoperative Adult Strabismus-20 scores improved numerically for each of the 4 Adult Strabismus-20 domains ( Figure ). Nevertheless, statistically significant improvement occurred only for the General Function domain, where median postoperative scores were significantly improved compared with preoperative scores ( P = .003, Figure ).