Health outcomes from cataract surgery using phacoemulsification
Date
1998-12
Authors
Borders, Julie A. Buddemeyer
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
With the success and popularity of treating cataract disorders, it is important to utilize health outcomes research to determine cataract surgery's effectiveness and benefits. Significant indicators that predict and measure success in cataract surgery need to be identified and verified. Statistical models can then be developed to predict successful outcomes. Better outcomes will enhance patient satisfaction and better utilize health resources.
The purpose of this study was to investigate patient visual acuity outcomes after the use of phacoemulsification during cataract surgery. The visual acuity outcomes from 140 consecutive surgeries were divided into three groups of improvement. One hundred-five eyes obtained better than or equal to 20/40 (0.500) visual acuity. The 20/40 visual acuity level is used in most states for an unrestricted driving license Masket (as cited in Lee, et al., 1993). Most literature uses the cutoff of 20/40 or better to designate successful visual outcomes following surgery (Lee, et al, 1993, p. 3). Three eyes showed worse results from pre-operative best corrected glare visual acuity to post-operative best corrected visual acuity. Thirty-two eyes showed some improvement from pre-operative best corrected glare visual acuity to post-operative best corrected visual acuity.
Because patient visual acuity outcomes for the group that became worse was so few in number, only three, the cell size made analysis difficult. For this reason only one research question could be explored. This was:
- What variables best predict those individuals likely to achieve 20/40 or better vision compared to those individuals who do not achieve the 20/40 standard?
This question is found in the literature (Schein, Cassard, and Javitt, 1995).
It should be noted that although an individual may not achieve the 20/40 or better level, they may still experience a substantial improvement in vision such that it improves their ability to function in daily life. An example of this would be an individual whose preoperative best corrected vision was 20/200 and best post-operative corrected vision is 20/60. Although that individual may not be able to see well enough to drive, she can once again see well enough to enjoy watching television, read, and move around more safely with less risk of falling. The goal of improving visual functioning, even if the visual acuity did not obtain the 20/40 level, could still be justified to managed care companies and other health care payors. Knowing the predictor variables for this group is important. It is also important to determine the predictor variables for the group of individuals for whom their visual acuity becomes worse after surgery. Knowledge of the predictor variables for these groups, would assist the ophthalmologist in determining who the best candidates are for cataract surgery. It would also improve the use of health care resources.
In the process of answering the research question, there were some additional issues to be explored which included:
- What are the significant indicators prior to receiving cataract surgery (demographic - age, race sex; physiologic - age, diabetic, hypertensive, degree of nuclear sclerosis; and/or visual comorbidities - age-related macular degeneration (AMD), glaucoma, diabetic retinopathy, high myopia) that may help predict the patient's ultimate visual acuity outcome?,
- Do significant differences exist in visual acuity outcomes between various populations, i.e. males and females, ethnic groups, diabetics and nondiabetics, hypertensives and nonhypertensives, "high" myopics and others who are not "high" myopics?,
- Are there any interactions occurring between predictor variables?,
- Are there any confounding factors, for example, high incidence rate of diabetes among Hispanics?
- If there are confounding factors, how do those factors influence cataract surgery outcomes?,
- Is there a significant difference in the mean average amount of phacoemulsification power required to pulverize the lens for each degree of nuclear sclerosis (hardness in the lens)?,
- Is there a significant difference in the mean average amount of phacoemulsification power required to pulverize the lens and those individuals who have post-operative complications and for those individuals who do not have post-operative complications?,
- Is there a significant association between pre-operative variables and post-operative complications?, and
- Is there a significant association between post-operative complications and ultimate visual outcome?
Research was conducted utilizing 140 ophthalmological medical records made accessible to the Southwest Texas State University Department of Health Services and Research. Data were collected and analyzed utilizing statistical modeling technique of logistic regression analysis. Using the results of the statistical modeling, an assessment was made regarding the most significant clinical indicators for cataract surgery.
The author reviewed 140 medical records and abstracted information to a data collection form (Appendix B). Variables on the questionnaire included demographic information, pre-existing medical conditions, ocular comorbidities, pre-operative and postoperative refractive measurements, and any post-operative complications. This information was entered into an Excel spreadsheet and analyzed in Excel and SPSS. The data were analyzed by descriptive analysis, univariate and comparative t-tests of sub-group means, chi-square analyses, linear regression and logistic regression.
Chi-square analyses indicated a significant association between group of improvement and age, and post-operative complications. Within this sample these factors tended to predict potential poor clinical visual outcome.
In answer to the issue of "Is there a significant difference in the mean average amount of phacoemulsification power required to pulverize the lens for each degree of nuclear sclerosis (hardness in the lens)?", analysis of variance with the least squares differences (LSD) test indicated individuals with nuclear sclerosis of 4+ degrees should be grouped with individuals of 5+ degrees because there is no significant difference in means of average phacoemulsification power used between these two groups. Individuals with 0 to 3+ degrees of nuclear sclerosis should be grouped together because there is no significant difference in means for average phacoemulsification power used between those four groups. Nuclear sclerosis was recoded and chi-square analysis showed a significant association for both nuclear sclerosis and operative complications (p <.01) and nuclear sclerosis and post-operative complications (p < .05). comparative t-tests of sub-group means indicated there was a significant difference (p< .05) in mean average phacoemulsification power for those individuals who had operative complications (mean was 1.72; CI 0.46,2.99) compared to those individuals who did not have operative complications (mean was 0.86; CI 0.66, 1.06). There was also a significant difference in mean average phacoemulsification power (p<.01) needed for those individuals who did not experience post operative complications (mean was 0.83; CI 0.63,1.02) compared with those individuals who did experience post-operative complications (mean was 1.67; CI 0.71, 2.64).
To answer the research question, the dependent outcome variable was coded dichotomously into zeros and ones. A logistic regression model was used to answer the question,
- "What are the predictor variables for those persons who achieve 20/40 or better vision compared to those individuals who do not achieve the 20/40 standard?"
Individuals who achieved the 20/40 or better best corrected visual acuity level were coded as "l". Individuals who did not achieve this level were coded as "0". The logistic regression model indicated that an interaction of age with age-related macular degeneration (AMD), the interaction of nuclear sclerosis with average phacoemulsification power, and post-operative complications were the best predictors of achieving or not achieving 20/40 or better best corrected visual acuity outcome.
Description
Keywords
cataract surgery, phacoemulsification, outcome assessment, medical care
Citation
Borders, J. A. B. (1998). Health outcomes from cataract surgery using phacoemulsification (Unpublished thesis). Southwest Texas State University, San Marcos, Texas.