Preoperative Expectations and Postoperative Outcomes of Visual Functioning among Cataract Patients in Urban Southern China

Purpose To investigate the relationship between preoperative expectations and actual postoperative outcomes of visual function (VF) among patients undergoing first eye cataract surgery. Methods A longitudinal study of 182 patients from hospitals in urban Southern China were surveyed prior to surgery and 3 month after cataract surgery regarding their preoperative, expected postoperative and actual postoperative VF for each of the items on the Catquest-9SF and their satisfaction with cataract surgery. In addition, detailed clinical data were collected preoperatively and postoperatively. Results The majority of cataract patients in urban Southern China had high expectations for VF outcomes after cataract surgery and in most cases postoperative outcomes achieved the expected level of improvement. The mean (standard deviation, SD) preoperative Catquest-9SF score was 15.7 (5.86) and the mean (SD) expected postoperative score was 26.3 (2.93). The discrepancy between actual and expected improvement was significantly correlated with patients’ health literacy, presence of systemic and ocular comorbidity, preoperative visual acuity of the surgery eye, LOCS III nuclear opalescence and cortical cataract grading. Conclusion Cataract patients in urban Southern China had high expectations for surgery outcomes. Patients with low level of health literacy and the presence of systemic and ocular comorbidity may need a comprehensive counseling to decrease the discrepancy regarding expected and actual outcomes.


Results
The majority of cataract patients in urban Southern China had high expectations for VF outcomes after cataract surgery and in most cases postoperative outcomes achieved the expected level of improvement. The mean (standard deviation, SD) preoperative Catquest-9SF score was 15.7 (5.86) and the mean (SD) expected postoperative score was 26.3 (2.93). The discrepancy between actual and expected improvement was significantly correlated with patients' health literacy, presence of systemic and ocular comorbidity, preoperative visual acuity of the surgery eye, LOCS III nuclear opalescence and cortical cataract grading.

Introduction
Cataract remains the leading cause of blindness and visual impairment in China [1,2]. The growing elderly population and advancements in medical technology have led to an increasing demand for cataract surgery [3,4] which in turn substantially increases functional vision and vision-related quality of life [5][6][7]. Recent advances in the developments of safe, effective and simplified techniques in cataract surgery have increased the demand for surgery and visual outcome expectations [8,9]. Recent data showed the 5-year incidence of cataract surgery in urban Southern China has improved, due to improved health care access [1]. In urban China, most cataract patients (89.6%) were willing to pay for cataract surgery although the costs for cataract surgery were higher than the average monthly income [10]. The patient-reported outcome and patient's satisfaction, reflected by postoperative visual outcome achieving the preoperative expectation, has become an important indicator for the success of cataract surgery outcomes [9,11]. Nevertheless, little information about preoperative patient expectations and postoperative patient satisfaction regarding the outcomes of cataract surgery is available in China, in which access, costs and quality of health care are changing rapidly.
Given the important role of vision function (VF) in daily life in elderly population, we aim to investigate the relationship between patient expectations and actual outcomes of VF after cataract surgery. In order to describe patients' self-assessed outcome after ophthalmic interventions in a more comprehensive way, numerous questionnaires have been developed, and these include VF-12 [8], VF-14 [11] and VF-15 [12]. The nine-item short-form Rasch-scaled Catquest-9SF was reliable and valid in measuring visual disability outcomes of cataract surgery [13][14][15], and has been validated previously in China [16]. In this study, we used Catquest-9SF questionnaire for patients undergoing the first eye cataract surgery in urban Southern China.

Methods Patients
The study included patients age 18 years or older awaiting first eye cataract surgery at the Second Affiliated Hospital and Sun Yat-sen University, Guangzhou, Southern China between Jan 2014 and Jun 2014. All patients completed the Catquest-9SF questionnaire preoperatively and again approximately 3 months after cataract surgery and an additional questionnaire for information about patients' ophthalmic and medical histories, demographic characteristics, health literacy and computer skill. The clinical information was collected by the examining ophthalmologist. Written informed consent was obtained from all participants. The study adhered to the Declaration of Helsinki and was conducted after obtaining ethical approval by the Zhongshan Ophthalmic Center Institutional Review Board.

Catquest-9SF questionnaire
Details regarding the Chinese version of the Catquest-9SF questionnaire has been described previously [16]. The Catquest-9SF questionnaire consisted of 7 questions for performing dailylife activities and 2 global questions about difficulties in general and satisfaction with vision.
For the perceived difficulty levels, the response options were as follows: 1 = very great difficulty; 2 = great difficulty; 3 = some difficulty; 4 = no difficulty. For satisfaction with vision, the response options were as follows: 1 = very dissatisfied; 2 = rather dissatisfied; 3 = fairly satisfied; 4 = very satisfied. Lower scores generally indicate worse VF.
The method to define patient expectations regarding the outcome of cataract surgery on VF referred to the previous study [8]. If the patients, who reported any difficulty on the Catquest-9SF tasks, considered there were chance of an improvement after surgery, the patients were asked to estimate how much improvement she or he expected (a little, a moderate amount, a great deal or do not know). Based on these responses, we calculated an expected postoperative Catquest-9SF score for each patient. This was done by adjusting each preoperative item-specific score upward by the amount of expected postoperative improvement (a little = 1; a moderate amount = 2, a great deal = 3). If an expected score exceed 4, it would be recorded as 4. The summary score was calculated for 7 items.

Questionnaire for health literacy, computer skills and demographic variables
For health literacy and computer skills measured by an interviewer-administered questionnaire were described previously [17]. Health literacy and computer skills were assessed by asking 3 questions. For health literacy, summary scores of > 10 and 10 were classified as inadequate and adequate, respectively. For computer skills, patients who answered ''yes" to all questions were considered to have adequate computer skills. Otherwise, patients were classified as having inadequate computer skills.

Clinical examination
Each patient underwent comprehensive eye examinations before and 3 months after cataract surgery. In addition to completion of the Catquest-9SF questionnaire, visual acuity, Lens Opacities Classification System (LOCS) III grading of cataract, and the type of implanted intraocular lens were recorded for each eye.
As introduced previously [18,19], the LOCS III cataract severity was assessed under pupil dilation and graded for each of the 4 possible components: nuclear opalescence, nuclear color, cortical cataract, and posterior subcapsular cataract. For each item, higher grading scores indicate greater severity. The scale ranges from 0.1 (clear or colorless) to 5.9 (very opaque in cases of cortical and posterior subcapsular cataract) or 6.9 (very opaque in cases of opalescence and nuclear color). Patients underwent standard phacoemulsification cataract surgery and intraocular lens (IOL) implantation.

Statistical methods
Only the surgical eye was included in the whole analysis. The main outcomes of the study were discrepancy between actual and expected postoperative Catquest-9SF scores (actual-expected) and the satisfaction with the surgery. Baseline data for demographic, health, ocular and visual functional characteristics of the patients were presented as mean (SD) for continuous variables with normal distribution and frequency (percentage) for categorical variables. Data for preoperative difficulty, expected and actual postoperative improvement on the Catquest-9SF items by demographic, ocular and functional characteristics were presented as median (Inter quartile range, IQR). The Wilcoxon signed rank sum test was used for comparing the difference between actual and expected postoperative Catquest-9 SF score by each level of the variables. The Kruskal Wallis test for variables with more than two levels and Wilcoxon-Mann Whitney test for variables with two levels were used for comparing the overall difference across groups of the variables between actual and expected postoperative Catquest-9 SF score. Chi-square test or Fisher's exact test (one or more than one cell frequency < = 5) was used to compare the proportion of the patients achieving expected improvement between groups. Spearman's rank correlation coefficients were calculated for measuring the strength of the association between functional characteristics and preoperative, expected and actual postoperative Catquest-9SF scores. Linear regression and ordinal logistic regression analyses were performed for the discrepancy between actual and expected postoperative Catquest-9SF scores and the satisfaction with the surgery respectively. Variables with P<0.05 in the simple regressions with one predictor in the model were included in the multiple regression analysis with more than one predictors in the model. All statistical analyses were performed using a commercially available software package (Stata 13.1, StataCorp, College Station TX, USA).

Results
Among 215 eligible patients, 182 were enrolled. Among these, 174 (95.6%) completed the detail preoperative and postoperative data and examination in the study. Table 1 lists the  * The preoperative Catquest-9SF score was the total score of each preoperative item-specific score with range from 7 to 28. ** The expected postoperative Catquest-9SF score was calculated by adjusting each preoperative itemspecific score upward by the amount of expected postoperative improvement coded as 1 = a little improvement, 2 = moderate improvement and 3 = great improvement. The maximum expected postoperative score for each item was limited to 4 and the range of the total expected postoperative score was 7-28.  Tables 2 and 3 show the preoperative difficulty, expected and actual postoperative improvement on the Catquest-9SF items by demographics and ocular characteristics. The median expected and actual postoperative Catquest-9SF scores were high, regardless of patient characteristics such as age, gender, education, health literacy, computer skill and presence of any systemic disease ( Table 2) and ocular characteristics such as preoperative visual acuity of the surgery eye, presence of any other ocular comorbidity, LOCS III grading and the type of implanted IOL ( Table 3). For the most part, the median difference between actual and expected postoperative scores was significantly zero, indicating the expected improvement was achieved. In some subgroups, the median actual postoperative score was lower than the expected postoperative score. Patients with inadequate healthy literacy, diabetes, preoperative visual acuity worse than 6/120, any ocular comorbidity (diabetic retinopathy), mild LOCS III NO and C grading and the implantation of aspherical IOL had a significantly larger difference between actual and expected postoperative Catquest-9SF scores. Table 4 shows the preoperative, expected and actual postoperative Catquest-9SF scores by the level of patient reported trouble and satisfaction with vision. Patients with better vision had higher preoperative VF scores (Spearman correlation, 0.53; P<0.001), higher expected postoperative scores (Spearman correlation, 0.23; P = 0.002), higher actual postoperative scores (Spearman correlation, 0.68; P<0.001) and larger difference between actual and expected postoperative scores (Spearman correlation, -0.26; P<0.001). The level of satisfaction with vision was significantly correlated with the preoperative score (Spearman correlation, 0.21; P = 0.005) and actual postoperative score (Spearman correlation, 0.65; P<0.001), but not correlated with the expected postoperative score and the difference between actual and expected postoperative scores. Table 5 shows the expected and actual postoperative improvement in specific items of the Catquest-9SF. The percentage of patients who had difficulty in each activity before surgery ranged from 81.0% to 94.8%. The portion of patients who expected to improve in activities with difficulty before surgery ranged from 82.1% to 95.8%. The proportion of patients who reported improvement after surgery among those with difficulty before surgery ranged from 86.1% to 91.8%. The ratio of patients who achieved expected level of improvement ranged from 70.1% to 81.3% for each activity. The portion of preoperative difficulty tended to be higher in near vision dependent activities, e.g., ability to read newspaper, read prices when shopping and do needlework. And the ratio of who achieved expected level of improvement tended to be lower, among these activities for which the near vision is required. Table 6 shows linear regression analysis of potential predictors of the difference between actual and expected postoperative overall Catquest-9SF scores. Simple linear regression analysis showed adequate health literacy (P = 0.04), presence of systemic comorbidity (diabetes)  (P = 0.009), any ocular comorbidity other than cataract (P<0.001), preoperative visual acuity of the surgery eye ( 6/120) (P = 0.008) and LOCS III grading C score (>4) (P = 0.04) were the significant predictors. The multiple analysis also showed presence of systemic comorbidity Table 3. (diabetes) (P = 0.04) and any ocular comorbidity other than cataract (P = 0.03) were the significant predictors. Patients with adequate health literacy had proper expectation for the postoperative VF outcomes which could be fulfilled. However, patients with poor prognostic   indicators (such as combined systemic or ocular diseases, poor preoperative visual acuity and severe degree of cataract) had larger difference between actual and expected improvement of VF. Table 7 shows the ordinal logistic regression analysis of potential predictors of the satisfaction with the surgery. The simple regression analysis showed patient satisfaction with the surgery was significantly associated with that presence of systemic comorbidity (hypertension) (P = 0.004), any ocular comorbidity other than cataract (P<0.001), LOCS III grading NO score (>3 to 5) (P = 0.03), LOCS III grading C score (>4) (P = 0.04) and the type of implanted IOL (P = 0.008). In the multiple analysis, presence of any ocular comorbidity (P = 0.007) appeared the major factor influencing patients' satisfaction with the cataract surgery.

Discussion
Using the Catquest-9SF questionnaire, we showed that the postoperative VF outcomes after cataract surgery achieved the expected level of improvement in the majority of cataract patients in urban Southern China. Our data showed that the discrepancy between actual and expected improvement was significantly associated with patient health literacy, presence of systemic and ocular comorbidity, preoperative visual acuity of the surgery eye and the degree of lens opacity, which was also associated with patient satisfaction with cataract surgery. Our findings are consistent with those from previous studies showing that patients' preoperative expectations were high, irrespective of the demographic or ocular characteristics. The achievement of expectations, which is considered as the discrepancy between actual and expected improvement, varied in different studies [8,9,11,12]. Tielsch et al. in 1991 showed that 61% of cataract patients (a total of 552 included) in the United States achieved or surpassed their expected level of postoperative VF at 4-month after cataract surgery using the VF-12 [8]. Pager in Australia showed 66% of patients (a total of 121 included) failed to equal or exceed their expectations at 1-month after cataract surgery using the VF-14 and the expectation-outcome discrepancy correlated significantly with satisfaction [11]. Addisu et al. used the VF-15 questionnaire to find that the actual outcome of cataract patients in Ethiopia was marginally better than expected improvement and the degree of improvement was not significantly related to patient satisfaction at all [12]. We have shown in an earlier study, that Chinese elderly patients with cataract had inadequate health literacy and computer skills, which were significantly associated with poor patientphysician communication and increase the risk of patient misunderstanding and mismanaging the ocular conditions [17]. Previous studies also suggested that improving patients' perceived level of understanding was an important factor to maximize patient satisfaction with cataract surgery [11,12]. Patients with adequate health literacy presumably had a full understanding of potential benefits and risks of cataract surgery [17]. It is therefore not surprising that adequate health literacy was a significantly positive predictor of the discrepancy between actual and expected postoperative improvement of VF. However, adequate computer skill was not correlated with patient satisfaction, and this could be attributed to the fact that most of our patients were more than 60 years old and the majority of them had very limited computer skills, leading to a lack of statistical power for identifying potential associations.
The presence of ocular comorbidity is known to be associated with poor patient-reported outcome [8,9,20]. Coexisting ocular comorbidity such as glaucoma, AMD, DR and others may influence patients' vision, leading to preoperative visual impairment and increasing the risk of surgical complication. For example, studies showed a significant association between primary angle-closure glaucoma eyes and a large deviation from the target refraction after cataract surgery, which retard the outcome of the surgery [20,21]. Stock et al. demonstrated that cataract surgery on eyes with AMD offers an increase in functional vision. However, when compared with patients without retinal pathology, AMD patients with poor preoperative visual acuity have worse visual outcomes [22]. In addition, we found that any systemic comorbidity (hypertension and diabetes) was significant correlated with the patient-reported outcome and satisfaction.
In our study, more than 50% of cataract patients had an extremely poor visual acuity before surgery and the majority of the eyes were graded less than 4.0 in LOCS III grading of nuclear opalescence and cortical cataract. LOCS III is a widely used cataract grading system, which reflects the type and severity of age-related cataract [23]. Pan's study showed that best corrected visual acuity (BCVA) was significantly correlated with the LOCS III nuclear opalescence grading but not LOCS III cortical cataract grading; and VF-14 score was significantly correlated with the BCVA and LOCS III nuclear opalescence grading of the better eye [24]. In our study, the poor surgery outcomes were significantly correlated with the low baseline VA and LOCS III cortical cataract grading (>4, the severe stage) of the surgery eye. In patients with early to moderate stages of cortical cataract, the opacities were located mainly at the periphery and the pupil area of the cortex was relatively clear. In patients with severe stage of cortical cataract, the pupil area of the cortex was opacified and the VF was obviously impaired. There is no significant correlation between patient-reported outcomes with the type of implanted IOL. However, patients with implantation of aspherical IOL are more likely satisfied with the surgery.
Strengths of our study included its prospective design and the high response rate. The data collected covered most of the major factors influencing the patient-reported outcome and patient's satisfaction with the cataract surgery. However, there were a few limitations. First, the patients were recruited from only two tertiary referral hospitals in Guangzhou and did not reflect the general cataract patients in urban southern China. Second, preoperative expected data were compared with postoperative actual data from patients receiving first eye cataract surgery. Most of patients need bilateral cataract surgery. A further study involving second eye surgery is required to validate our results.
In conclusion, cataract patients in urban Southern China have high expectations for surgery outcomes. Patients with low level of health literacy and the presence of systemic and ocular comorbidity may need a comprehensive counseling to decrease the discrepancy regarding expected and actual outcomes and to improve patients' satisfaction.