Accuracy of intraocular lens power calculation formulae after laser refractive surgery in myopic eyes: a meta-analysis

Author: Li H, Nan L, Li J, Song H.

Geographical coverage: No information provided on countries in which included studies were undertaken.

Sector: Biomedical

Sub-sector: Treatment

Equity focus: None

Study population: Patients who had laser-assisted in-situ keratomileusis (LASIK), photorefractive keratectomy (PRK) or laser-assisted subepithelial keratomileusis (LASEK) for myopia and subsequent uneventful cataract surgery.

Review type: Other review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: Patients who have undergone corneal excimer laser surgery are now confronting the prospect of cataract surgery as they age. Accurately determining the power of the intraocular lens (IOL) in eyes post-refractive surgery poses a significant challenge for many ophthalmologists. Over the years, numerous methods have been proposed to enhance the precision of IOL power calculation in patients who have had corneal refractive surgery. The clinical history method was once deemed the gold standard. However, there are instances where historical data is either unavailable or unreliable. Several formulas that do not rely on historical data have been developed, including the Barrett True-K no history, Double-K method, Haigis-L, OCT formula, Shammas-PL, Wang-Koch-Maloney (W-K-M), and various IOL calculators, among others. While these formulas offer greater accuracy than traditional formulas and the historical data method, their predictability varies across studies. The question of which formula is best for calculating IOL power in eyes post-laser refractive surgery remains a topic of debate.

Objectives: To compare the accuracy of intraocular lens power calculation formulae after laser refractive surgery in myopic eyes.

Main findings:

From an initial pool of 1,144 articles, 16 were ultimately selected for analysis in this meta-study, involving 1,167 eyes. The studies primarily focused on patients with mono-focal IOL implants.

The meta-analysis used seven common methods for IOL power calculation, with ASCRS average, Barrett True-K no history, and OCT formula showing significantly higher accuracy. ASCRS average yielded a significantly higher percentage of refractive prediction error within ±0.5 D than Haigis-L, Shammas-PL and Wang-Koch-Maloney (p = 0.009, 0.01, 0.008, respectively). Barrett True-K no history also yielded a significantly higher percentage of refractive prediction error within ±0.5 D than Shammas-PL and Wang-Koch-Maloney (p = 0.01, p <0.0001, respectively), and a similar result was found when comparing OCT formula with Haigis-L and Shammas-PL (p = 0.03, p = 0.01, respectively).

The sensitivity analysis showed that by omitting Cho 2018, authors found a significant decreased to 0% in the comparison of the percentage of refractive prediction error within ±1.0 D between Haigis-L and Shammas-PL and between Haigis-L and W-K-M. Cho et al (2018) did not show details of the IOL constant optimisation procedure and did not report whether the keratometry was measured from the IOL Master. After omitting this study, the I² value decreased.

Despite of the limitations of included studies in the review, the authors recommend the use of ASCRS average or Barrett True-K no history for IOL power calculation after myopic laser refractive surgery, with the OCT formula as a good alternative. Further studies are needed for hyperopic eyes.

Methodology:

The studies included in the analysis met the following criteria: (1) they involved patients who had undergone LASIK, PRK, or LASEK for myopia and subsequently had uneventful cataract surgery; (2) they used at least two types of IOL power calculation formulas, including ASCRS average, Barrett True-K no history, Double-K SRK/T, Haigis-L, OCT formula, Shammas-PL, and W-K-M; (3) they measured optical biometry using PCI (IOL Master); (4) they optimized IOL constants. Studies were excluded if: (1) they involved patients who had undergone hyperopic refractive surgery or radial keratotomy surgery; (2) they did not provide data on the percentage of refractive prediction error within ±0.5 D and ±1.0 D; (3) they involved eyes that did not have in-the-bag fixed IOL implantation or had undergone another ocular surgery. Additionally, the Intraoperative refractive biometry, Shammas-PHL, and Olsen T formulas were excluded due to their limited clinical use.

Two independent researchers conducted a comprehensive search of the PubMed, EMBASE, Web of Science, and Cochrane Library databases. They sought relevant studies published from January 1, 2009, to August 11, 2019, using various search terms mentioned in the articles. All potential studies were considered for review, irrespective of the primary outcome or language. The titles and abstracts of all the searched studies were evaluated by two authors, who also manually searched the reference lists of all eligible articles. Two authors independently extracted and compared the data, using a modified QUADAS-2 tool checklist to assess the quality of the evidence.

The study used seven different formulas to calculate the IOL power in eyes after myopic laser refractive surgery. The primary outcomes were the percentages of refractive prediction error within ±0.5 D and ±1.0 D, with a higher percentage indicating greater precision. The refractive prediction error was calculated by subtracting the predicted spherical equivalent from the actual postoperative spherical equivalent. Pooled estimates of the odds ratio (OR) were calculated for categorical outcomes using a fixed-effects model. The I² statistic was used to quantify heterogeneity across studies. Sensitivity and subgroup analyses were conducted when the I² value was greater than 50%. Funnel plots were used to evaluate publication bias and small-study effect. Data pooling was done using Review Manager, with a p-value less than 0.05 considered statistically significant.

Applicability/external validity:

The authors of this review do not explicitly assess the applicability or external validity of their findings. They acknowledge that their meta-analysis has several limitations, such as the variability of patient characteristics, IOL types and single-center analysis, the use of PCI for optical biometry, and the exclusion of hyperopic refractive surgery eyes and other recent formulae.

Geographic focus:

The authors did not explicitly discuss how their findings may vary in different geographic regions.

Summary of quality assessment:

The procedures employed for identifying, including, and critically evaluating studies were fairly solid, but there’s no indication that unpublished materials were considered for inclusion or that relevant experts were consulted. Furthermore, there’s no clear rationale provided for choosing 2009 as the starting point for the search. The data analysis methods were generally sound, but there’s no evidence that the analysis took into account the quality of the included studies or other factors that could have influenced the results. This may not have been feasible given the limited number of studies and the wide range of methods involved. Due to these factors, we have medium confidence in the review’s findings.