Comparing the Accuracy of the New-Generation Intraocular Lens Power Calculation Formulae in Axial Myopic Eyes: A Meta-Analysis

Authors: Li H, Ye Z, Luo Y, Li Z.

Geographical coverage: Australia, China, India, Ukraine, Colombia, the United States, Spain, and Germany

Sector: Intraocular lens

Sub-sector: Accuracy

Equity focus: Not reported

Study population: Patients with cataract

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

The global rise in myopia has increased the number of patients with elongated axial lengths. Cataract surgery in highly myopic eyes poses a significant challenge for accurate intraocular lens (IOL) power calculation. In axial myopic eyes, errors in axial length (AL) measurement, anterior chamber depth (ACD), and particularly effective lens position (ELP) estimation are the main contributors to refractive inaccuracies. Traditional third- and fourth-generation IOL power calculation formulae (such as SRK/T, Hoffer Q, Holladay I/II, and Haigis) have shown variable results in these eyes, often leading to postoperative hyperopic shifts. New-generation formulae, such as Barrett Universal II (BUII) and Olsen, and artificial intelligence-based formulae like Kane and Hill-RBF, have emerged, integrating more biometric variables and advanced algorithms to improve prediction accuracy. Despite these advances, the optimal formula for axial myopia remains undetermined, requiring further comparative analyses to guide clinical decision-making.

Objective:

To compare the accuracy of new-generation IOL power calculation formulae in axial myopic eyes.

Main findings:

The review included 15 studies encompassing a total of 2,273 eyes. Regarding methodological quality, one study had a high risk of bias in patient selection, and seven had an unclear risk of bias. In terms of reference standard as well as flow and timing, six studies performed postoperative subjective refraction, whereas three did not describe their refraction method. For the index test, only two studies failed to define the outcome measures.

The findings revealed that, for achieving a refractive prediction error within ±0.5 dioptres (D) of target, Barrett Universal II (BUII) was significantly more accurate than all traditional formulae (SRK/T, Hoffer Q, Holladay I/II, Haigis)1. In addition, BUII performed similarly to other new-generation or AI-based formulae (Olsen, Kane, Hill-RBF, Ladas Super Formula (LSF), Emmetropia Verifying Optical (EVO), T2)1. For refractive outcomes within ±1.0 D, BUII also significantly outperformed all traditional formulae (SRK/T, Hoffer Q, Holladay I, Holladay II, Haigis, T2)1, and it performed equally well as the other new-generation formulae (Olsen, EVO, Kane, Hill-RBF, LSF)1.

In summary, among the new-generation IOL power calculation formulae evaluated, BUII demonstrated the most consistent and statistically significant superiority over all third- and fourth-generation formulae in axial myopic eyes (based on the proportions of eyes within ±0.5 D and ±1.0 D of predicted refraction). While other new-generation formulae (Olsen, Kane, Hill-RBF, EVO, LSF) also showed high accuracy, their advantages over the older formulae were not always statistically significant. Therefore, BUII stands out as the most accurate and reliable formula for IOL power calculation in this patient population.

Methodology:

Searches were conducted in PubMed, Embase, Web of Science, and the Cochrane Library for studies published in English between 11 April 2011 and 11 April 2021. Studies were included if they involved eyes with AL > 24.5 mm and compared at least two different IOL power calculation formulae. Reference lists of all eligible studies were also scanned for additional publications. Two reviewers independently screened the articles, extracted data, and critically appraised study quality using a modified QUADAS-2 checklist. Disagreements were resolved by consultation with a third reviewer. The data were synthesised using fixed- or random-effects meta-analyses, depending on heterogeneity (assessed with the I² statistic). Sensitivity and subgroup analyses were performed when I² exceeded 50%. Funnel plots were used to evaluate publication bias and small-study effects.

Applicability/external validity:

The review noted that some included studies were retrospective with small sample sizes, potentially causing selection bias. Additionally, the axial length ranges differed among studies (even though all eyes had AL > 24.5 mm), which could have had a minor effect on the results. Moreover, only one study included the EVO, Kane, and LSF formulae; therefore, conclusions regarding these three formulae have limited applicability. The authors emphasised the need for further well-designed studies to validate and extend these findings for IOL power calculation in axial myopic eyes.

Geographic focus:

No geographical limits were applied. The included studies were conducted in a variety of countries, including Australia, China, India, Ukraine, Colombia, the United States, Spain, and Germany.

Summary of quality assessment:

Overall, there is medium confidence in the review’s conclusions. The literature search was comprehensive, and inclusion/exclusion criteria were clearly defined. Two reviewers independently conducted study selection, data extraction, and quality assessment using established tools, with disagreements resolved by a third reviewer. Characteristics of the included studies were well documented, and the meta-analyses were appropriately performed with heterogeneity addressed. However, the search was limited to English-language publications, the review did not provide a list of excluded studies, and the results were not stratified by risk-of-bias status. These limitations suggest that the findings should be interpreted with some caution.