Accuracy of intraocular lens power calculation formulae in short eyes: A systematic review and meta-analysis

Author: Shrivastava AK, Nayak S, Mahobia A, Anto M, Pandey P.

Geographical coverage: United Kingdom, the United States, India, South Korea, Australia, Thailand, Germany, and China

Sector: Cataract surgery

Sub-sector: Intraocular lens

Equity focus: Not reported

Study population: Patients with cataract

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

Cataract surgery is one of the most performed ophthalmic procedures throughout the world. The accurate intraocular lens (IOL) power calculation is key to achieving optimal visual outcomes. Optical biometry has improved precision and minimised errors associated with ultrasound biometry. While modern formulae perform well in eyes with normal axial length, short eyes (AL <22 mm) present greater challenges due to anatomical differences. Many IOL power formulae exist, but none are consistently accurate for all cases. Previous studies have shown variable outcomes in short eyes, which complicates formula selection. However, updated standards now emphasise reporting median absolute error (MedAE) and prediction errors within ±0.50 D and ±1.0 D to evaluate accuracy more precisely.

Objective:

To evaluate the accuracy of intraocular lens power calculation formulae in short eyes.

Main findings:

The review included 15 studies (2,395 eyes with AL ranging from 18.77 to 21.99 mm) in the meta-analysis to evaluate the accuracy of 12 intraocular lens power calculation formulae. The sample sizes in the included studies ranged from 15 to 608 eyes. The included studies were conducted in the United Kingdom (n=4), the United States (n=2), India (n=2), South Korea (n=1), Australia (n=2), Thailand (n=1), Germany (n=1), and China (n=2).

The results showed that Barrett Universal II had the lowest mean absolute error (MAE) and median absolute error (MedAE), suggesting it may be the most accurate formula for short eyes. However, statistical comparisons revealed no significant differences between Barrett Universal II and other formulae, including Haigis, Hoffer Q, and SRK/T. Similarly, Holladay 1 had the highest percentage of eyes within ±0.50 D of the target refraction, while Hill-RBF had the highest percentage of eyes within ±1.0 D. Despite these trends, none of the formulae demonstrated statistically superior performance over the others in any of the comparisons.

In conclusion, while Barrett Universal II, Holladay 1, and Hill-RBF showed favourable outcomes in specific metrics, no single formula was statistically superior for IOL power calculation in short eyes.

Methodology:

The searches were conducted in PubMed, Embase, the Cochrane Library, ScienceDirect, Web of Science, and Scopus to identify studies involving adult patients with axial length <22 mm who underwent optical biometry or uneventful phacoemulsification surgery. Studies that reported MAE, MedAE, or the percentage of eyes with prediction error (PE) within ±0.50 D or ±1.0 D of the target refraction, and were published in English in the past 21 years (up to July 2021), were included in this review. Reference lists of relevant reviews were also carefully scanned to identify additional relevant studies.

Two reviewers independently extracted the relevant data. Any disagreements between the reviewers were resolved through discussion or by involving a third reviewer. The methodological quality of the included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies–2 (QUADAS‑2) tool. The findings were synthesised using a fixed- or random-effects model meta-analysis based on heterogeneity. Heterogeneity was assessed using the Chi-square test and I² statistic. Sensitivity analysis was conducted to identify sources of heterogeneity, and publication bias was assessed using a funnel plot.

Applicability/external validity:

The review discussed the applicability of the findings, noting that the included studies originated from various geographic regions, which enhances the generalisability of the results.

Geographic focus:

The review did not apply any geographical limits. The included studies were conducted in the United Kingdom, the United States, India, South Korea, Australia, Thailand, Germany, and China.

Summary of quality assessment:

Overall, confidence in the conclusions drawn from this review is low. The literature search was comprehensive, with clearly defined inclusion and exclusion criteria. Study quality was assessed, and the results of these assessments were reported. Two reviewers independently extracted the data from the selected studies. Characteristics of the included studies were presented. The data were synthesised using meta-analysis, and heterogeneity was assessed using the Chi-square test and I² statistic. Subgroup and sensitivity analyses were also conducted. However, the search was limited to articles published in English. The review did not provide a list of excluded studies, and the findings were not reported according to risk of bias. Additionally, it was unclear how many reviewers screened the studies.