REVIEW ON PHACO SURGERY

Formulas for Intraocular lens Calculation in Phacorefractive

Surgery of Patients with high Myopia. Which is the most
Accurate? Systematic Review and Meta-analysis of the Literature
Fórmulas para Cálculo de lente Intraocular en Cirugía Faco
Refractiva de Pacientes con Miopía alta ¿Cual es la más
Precisa? Revisión Sistemática y Metanálisis de la Literatura
Andrea Armería-Díaz de León1, Mauricio Pierdant-Pérez2, Cristhian B Camera-Miranda3, Richard D Tirado-Aguilar4,
Vicente Esparza-Villalpando5
Received on: 27 February 2023; Accepted on: 18 May 2023; Published on: 13 July 2023

A b s t r ac t
Introduction: Myopia is the most common ocular condition worldwide. Various methods exist for correcting high myopia, and our focus lies
on refractive extraction of the crystalline lens with intraocular lens implantation (IOL). The IOL calculation should be as accurate as possible,
therefore, a systematic review was performed to compare the error in the prediction of postoperative refraction among different formulas.
Objective: To carry out an evaluation of the literature found in the main electronic databases, where the results of phacorefractive surgery in
patients with high myopia are described based on the predictive error of the formulas used.
Methods: Search for information in specialized electronic information sources: PubMed, Trip Database, Cochrane Library and Science Direct.
Meta-analysis of the MAE (mean absolute error) and MNE (mean numerical error) results of the selected articles was performed to estimate
which formula is the most accurate for calculating the intraocular lens in eyes with high myopia.
Results: 10 articles were selected that met the eligibility and quality criteria for the systematic review, of which 7 articles were used to perform
the meta-analysis of single means of MAE and 6 articles for the meta-analysis of single means of MNE, the Barrett formula Universal II obtained
the lowest global values of
​​ predictive error in both meta-analyses.
Conclusions: The Barrett formula, as it obtains the most predictable results, is considered the standard formula for intraocular lens calculation
in patients with high myopia.
Keywords: High myopia, Intraocular lens, Phaco refractive surgery.

Resumen
Introduccion: Miopía es la condición ocular más frecuente a nivel mundial. Existen diferentes procedimientos para corregir la miopía alta, nos
enfocamos en la extracción refractiva del cristalino con implante de lente intraocular (LIO). El cálculo de LIO debe ser lo más preciso, por lo tanto,
se realizó una revisión sistemática para comparar el error en la predicción de refracción posoperatoria entre diferentes fórmulas.
Objetivo: Revisión de la literatura en diversas bases de datos electrónicas, que describan el error predictivo de las fórmulas empleadas en cirugía
faco refractiva en pacientes con miopía alta.
Metodos: Búsqueda de información en las fuentes de información electrónicas especializadas. Se realizó metanálisis de los resultados de MAE
(error absoluto medio) y MNE (error numérico medio) de los artículos seleccionados para estimar qué fórmula es la más exacta para calcular
el lente intraocular en ojos con miopía alta.
Resultados: Se seleccionaron 10 artículos que cumplieron los criterios de elegibilidad y calidad, de los cuales, se utilizaron 7 artículos para
realizar el metanálisis de medias únicas de MAE 0.33 (0,23; 0,42) para fórmula de Barrett Universal II y 6 artículos para el metanálisis de medias
únicas de MNE siendo de -0.09 (-0,13; -0,05) para la misma, ésta obtuvo los valores globales menores de error predictivo en ambos metanálisis.
Conclusiones: La fórmula de Barrett al obtener los resultados más predecibles, se considera la fórmula estándar para cálculo de lente intraocular
en pacientes con miopía alta.
Palabras clave: Cirugía faco refractiva, Miopía alta, Lente intraocular, Longitud axial.
Revista Mexicana de Oftalmología (ENG) (2023): 10.5005/rmo-11013-0020

Introduction 1,3
Department of Ophthalmology Hospital Central “Ignacio Morones
A refractive error arises from an anatomical condition of the eye in Prieto”, San Luis Potosi, S.L.P, Mexico
which there is a disparity between the axial length of the eye and 2,4,5
Department of Public Health and Medical Sciences, Faculty of
its optical power,1 causing objects’ images to not be focused on the Medicine San Luis Potosi, S.L.P, Mexico
retina. There are three main types of refractive errors: hyperopia Corresponding Author: Mauricio Pierdant Pérez, Department of
and myopia, which represent the so-called spherical errors, and Ophthalmology Hospital Central “Ignacio Morones Prieto”, San Luis
astigmatism, which involves optical asymmetry.1 Potosi, S.L.P, Mexico, e-mail: mauricio.pierdant@uaslp.mx

2604-1731 / © 2023 Sociedad Mexicana de Oftalmología. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/)
 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

In 2010, it was estimated that uncorrected refractive errors How to cite this article: Armería-Díaz de León A, Pierdant-Pérez
were the leading cause of distance vision impairment, affecting M, Camera-Miranda CB, Tirado-Aguilar RD, Esparza-Villalpando V.
108 million people, and the second leading cause of treatable Formulas for intraocular lens calculation in Phacorefractive Surgery
blindness worldwide.2 of patients with high myopia. Which is the most accurate? Systematic
Myopia is the most common ocular condition globally,3 with the Review and Meta-analysis of the Literature. Rev Mex OftalMOl. (ENG)
2023;97(2):40–54.
highest prevalence reported in Asians (18.5%), where it is considered
epidemic proportions2, followed by Hispanics (13.2%) and African Source of support: Nil
Americans (6.6%).3 Conflict of interest: None
Myopia can be defined as a condition characterized by abnormally
increased axial length of the eye (greater than 24 mm) and a spherical
equivalent greater than -0.5D. The definition of high myopia varies in The SRK/T formula is a well-known method with reported
different studies and has been standardized as a spherical equivalent evidence of its accuracy in cases of high myopia. Postoperative
of -5.00D or greater or an axial length greater than 26 mm.4 anterior chamber depth and IOL position are predicted by SRK/T.
Various procedures have been developed to correct high As a result, this formula applies a correction factor for eyes with
myopia, including contact lenses, laser refractive surgery (Laser an axial length greater than 24.2 mm, favoring a more accurate
in situ keratomileusis (LASIK) and advanced surface ablation estimation of the anterior chamber depth in long eyes.12
techniques), phakic intraocular lenses, and refractive lens extraction The Holladay 1 formula has also been successful in emmetropic
with intraocular lens implantation (IOL).5 and myopic eyes. 12
Although laser refractive surgery is the most commonly Barrett Universal II uses a theoretical eye model in which the
performed refractive surgery worldwide, it has limitations regarding anterior chamber depth is related to axial length and keratometry.
visual quality, dry eye, and the range of achievable refractive The formula was described as universal because it aims to work
correction. 6 It is also associated with corneal complications for different types of lenses and for eyes with short, medium, and
such as ectasia, flap loss, or infection. In order to prevent these long axial length.15
complications, the approach of substituting the natural crystalline Standard formulas (third generation) tend to select IOLs with
lens with an intraocular lens has been reintroduced.5 insufficient power, leaving patients with postoperative hyperopia,16
The concept of extracting the crystalline lens from patients which is not the desired outcome in practice. In patients with high
diagnosed with high myopia for the purpose of refractive correction myopia, the recommended postoperative refractive outcome is
is not a recent innovation; historical records indicate its origins in -2.0 to -3.0 D, primarily because most high myopic patients prefer
the 18th century.7 better near vision than distance vision.17 Some exceptions may
In 1776, Abbe Desmonceaux was the pioneering figure who include young patients who have had adequate lens correction
initially proposed the idea of lens extraction as a treatment for and prefer achieving emmetropia postoperatively.17
myopia. Subsequently, in 1890, Fukala published his work on the Fourth-generation formulas also take into consideration the
treatment of high myopia through aphakia.8 patient’s age, preoperative refraction, anterior chamber depth,
Refractive lens exchange is, by definition, a surgery in which a lens thickness, and white-to-white distance.14
cataract or clear crystalline lens is replaced with an intraocular lens Other formulas have shown promising results, such as the Kane
in cases of high ametropia.8 formula, EVO (Emmetropia Verifying Optical), and Panacea formula.18
The main complication is the rupture of the posterior Formulas for optimizing axial length in patients with axial
capsule of the lens, with an incidence ranging from 0.50% to myopia have been proposed. Liu et al. developed the Wang-Koch
1%.9 Furthermore, the removal of the crystalline lens causes formula for adjusting axial length for use with the Haigis, Hoffer
volumetric changes in the eye, leading to vitreous degeneration Q, Holladay 1, SRK/T, and Holladay 2 formulas. It has been found
and movement that may predispose to rhegmatogenous to produce myopic results, leading to the development of the
retinal detachment.10 Vukicevic et al. Reported an incidence of modified Wang-Koch formula, which uses the lens constants from
subclinical cystoid macular edema following uncomplicated the User Group for Laser Interference Biometry (ULIB).19
phacoemulsification of 5%. 5 Partial coherence interferometry optical biometry allows for
Modern intraocular lens (IOL) calculation formulas provide more accurate measurements, even in the presence of staphyloma.13
accurate results for eyes with an axial length (AL) of 22-25 mm It shows advantages in accuracy and reproducibility of axial length
with today’s technology. However, there is controversy regarding determination, and is more comfortable for both the patient and
the accuracy of IOL11 power calculation formulas for eyes with the investigator.18 As a result, it has become a standard criterion
high axial length. These eyes often have a deep anterior chamber, for eyes undergoing crystalline extraction and intraocular lens
which typically leads to refractive surprise when calculating IOL implantation.18 Its feasibility is limited in cases of dense cataracts,
power12. retinal detachment, and fixation issues.18
Inadequate measurement of preoperative AL has been reported Ultrasound biometry encounters challenges in specific clinical
as the main reason for refractive error in high axial myopia. The conditions such as deformities of the eyeball, myopic staphyloma,
primary cause of refractive errors in high axial myopia has been eccentric fixation, or the presence of silicone in the posterior
attributed to inaccurate preoperative measurements of axial segment.18
length.13 Accurate calculation of IOL power is of utmost importance The improved efficacy, predictability, and safety of modern
to meet postoperative refractive expectations.11 phacoemulsification, along with more accurate methods for
The initial formulation of intraocular lens calculation formulas measuring corneal curvature, axial length, better IOL calculation
was aimed at achieving the intended postoperative refraction in formulas, and improved surgical techniques, have led to the
cases of pseudophakia.14 resurgence of crystalline extraction as a modality for correcting high

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023) 41

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

myopia.8 It is not just a matter of extraction but also a refractive specialized information sources such as PubMed, Trip Database,
procedure.20 Cochrane Library, and Science Direct. Keywords, MeSH terms, and
The objective of conducting a systematic review, following the finally the described DeCS terms (Table 1) were utilized.
guidelines outlined in the Cochrane manual, is to comprehensively Initially, the articles were selected based on their titles by one
gather all relevant empirical evidence that meets the predetermined of the authors. Following that, the investigators independently
eligibility criteria in order to address a specific research question. reviewed the abstracts of the selected articles.
The purpose of conducting a systematic review, based on the Afterward, one of the authors thoroughly examined the full-
guidelines provided by the Cochrane Manual, is to gather all the text versions of potentially relevant studies. To make an informed
empirical evidence that meets the pre-specified eligibility criteria decision on their inclusion in the analysis, a quality scale was
in order to answer the specific research question.20 employed. The original articles were evaluated using the OPMER
Deviation from the desired refraction is a prevalent reason for scale to assess the quality of evidence.
secondary interventions that may be required after intraocular lens This scale was developed in the Department of Epidemiology
implantation.15,18 Improvement in vision represents the primary goal at the Autonomous University of San Luis Potosí and is based
of crystalline surgery; however, refractive outcome has become on the PRISMA criteria. It consists of a checklist that includes
increasingly relevant, especially in myopic patients, when clear five validity groups addressing the scientific robustness of the
lens exchange is used as a refractive surgery.18 Predictive error is original articles: Objective, Population, Methodology, Statistics,
calculated by subtracting the postoperative refraction from the and Results. 21
predicted refraction. The resulting numerical difference between The OPMER guideline comprises five sections, and each
the predicted and obtained values is known as the predictive error. section incorporates three essential determinants aimed
The Mean Numerical Error (MNE) is then computed by averaging at ensuring appropriate methodological functioning. Each
these numerical differences across all evaluated formulas. determinant is assigned a value of one point if it is present and
Additionally, the Mean Absolute Error (MAE) was calculated,28 zero points if it is absent. Additionally, each section contains a
representing the average of the predictive errors taken as absolute main determinant, which was weighted accordingly by a group of
values for each studied formula. This reflects the systematic bias of expert methodological peers. These weighted determinants have
the formula. A negative value indicates a tendency towards under a value of two points if complete, one point if incomplete, and zero
correction, resulting in a hyperopic outcome compared to the points if absent.21
predicted refraction, and vice versa.14 The Median Absolute Error The total maximum score that can be achieved by summing up
(MedAE) is the median of the MAE values. all the possible points from the OPMER guideline is 20 points. The
One of the reasons why myopic patients require cataract five main determinants contribute 10 points to the total score and
surgery at a younger age is mainly because nuclear cataracts constitute a methodological axis of quality on their own.
develop earlier in eyes with high myopia.17 In relation to the obtained score, repeatability and agreement
tests conducted with this instrument aim to determine whether an
article lacks methodological robustness if it receives a final score of
Objective less than 10 points. If the article obtains between 11 and 14 points,
The objective is to assess relevant literature from major databases, its methodological quality is in doubt, and the number of missing
analyzing outcomes of phaco refractive surgery in high myopia main determinants must be determined.21
patients. This evaluation focuses on the prediction error of various Lastly, if an article achieves a score of more than 15 endpoints,
intraocular lens calculation formulas to determine the formula with with at least 3 main determinants present, it is categorized as a
the lowest predictive error. methodologically well-structured medical article. The articles that
obtained more than 14 points were selected.
Methods
Search Strategy D ata b a s e s and I n f o r m at i o n S o u r c e s
A qualitative systematic review was conducted based on the The databases and metasearch engines used in this article were:
Preferred Reporting System Items for Systematic Reviews and PubMed, Science Direct, Cochrane Library and Trip Database.
Meta-Analyses (PRISMA) criteria. (((“Phacoemulsification”[Mesh] AND “Lens Implantation,
The advanced information search (BIS) was initiated in Intraocular”[Mesh]) AND “Myopia”[Mesh]) NOT “Myopia,
September 2021 and concluded in November 2021, using Degenerative”[Mesh]

Table 1:  Search strategy

KEYWORD DECS SYNONYMS MESH SYNONYMS
Facoemulsificación The same The same phacoemulsification The same
Intraocular lens implantation Intraocular lens The same Lens Implantation, Implantation, Intraocular Lens
implantation Intraocular Implantations, Intraocular Lens
Intraocular Lens Implantation
Intraocular Lens Implantations
Lens Implantations, Intraocular
Myopia Myopia Shortness of vision Myopia Myopias
Nearsightedness
Nearsightednesses

42 Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

(((“Phacoemulsification”[Mesh] OR “Refractive Lens Exchange”[T/ Finally, after reading the selected articles, if the OPMER score
AB] OR “Clear Lens Extraction”[T/AB] OR “Clear Lensectomy”[T/AB]) was below 14 points, the study was excluded (Fig. 1).
AND “Lens Implantation, Intraocular”[Mesh]) AND “Myopia”[Mesh]
OR “Nearsightedness”[MESH]) NOT “Myopia, Degenerative”[Mesh]
R e s u lts
L i m i ts A total of 1458 documents were identified through the search
strategy, which were screened by title and abstract, resulting in
In databases like PubMed, we used the “human studies” filter; no
the exclusion of 1359 articles. 99 articles that included the specified
filters were needed for the other meta-search engines.
descriptors were further analyzed. Using the Zotero bibliographic
manager, we identified 37 duplicate articles, and two researchers
S e l e c t i o n C r i t e r ia independently and blindly verified their duplication status.
Original studies published after 1994, conducted in humans of Articles were retrieved for the eligibility assessment, from which
any age, with a diagnosis of high myopia (spherical equivalent of studies with an OPMER score of less than 14 and those that did not
-5.00D or greater or axial length greater than 26 mm), and in which meet the inclusion criteria were eliminated.
transparent lens extraction and intraocular lens insertion were Ten original articles met the criteria and were selected as the most
performed as refractive treatment were included. important and reliable. These 10 studies included a total of 1094
patients, aged between 38 and 93 years. All of the included patients
E xc lu s i o n C r i t e r ia in these studies were diagnosed with high myopia. None of these
We excluded studies with titles unrelated to the main topic, patients experienced intraoperative complications or underwent
studies that were repeated across databases, studies related to any other procedures during surgery. Additionally, none of the
ophthalmologic procedures other than phacoemulsification (such patients had other diagnoses such as keratoconus, glaucoma, retinal
as extracapsular cataract extraction, corneal refractive surgery, detachment, or a history of corneal refractive surgery.
vitrectomy, cerclage, glaucoma surgery, etc.), studies involving Liu et  al. investigated the accuracy of intraocular lens (IOL)
patients with keratoconus, studies involving phakic intraocular power calculation in eyes of Chinese patients with high myopia
lens implantation, studies involving patients without a diagnosis using two new formulas (Barrett Universal II and Hill-RBF 2.0),
of high myopia, studies that did not mention the formula used three unmodified formulas (Haigis, Holladay 1, and SRK/T), and the
for intraocular lens calculation, and studies that did not provide original and modified versions of the Wang-Koch formula adjusted
information on the formula used for intraocular lens calculation. for axial length with the Holladay 1 and SRK/T formulas.19
Studies that did not report results of MAE (Mean Absolute Error) Haigis, Holladay 1, and SRK/T formulas resulted in hyperopic
or MNE (Mean Numerical Error) calculation were also excluded. numerical error (MNE) The formulas with modified Wang-Koch

Fig. 1:  BIS flow diagram

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023) 43

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

showed myopic error. The formulas with modified Wang-Koch were In the study conducted by Terzi et al., they evaluated the results
less aggressive in producing myopic outcomes compared to the using four modern formulas for IOL calculation: Holladay 2, Hoffer
original Wang-Koch formula, with an incidence of hyperopic results Q, SRK/T, and Haigis, in eyes with a length greater than 26.00 mm
similar to Barrett (28-36%), but lower than the other formulas.19 or less than 22 mm.14 Preoperatively, IOL calculation was performed
Barrett Universal II produced a significantly lower mean using the Holladay 2 formula, and retrospectively using Haigis,
absolute error (MAE) than Holladay 1, SRK/T, SRK/T w/k, and Hoffer Q, and SRK/T.14
modified SRK/T w/k. When using the ULIB constants, Barrett All formulas yielded a negative Mean Numerical Error (MNE),
Universal II and Hill-RBF formulas yielded a near-zero numerical and 77-95% had a refractive surprise towards hypermetropia. The
error (MNE). The Barrett, Hill-RBF, and original and modified Wang- absolute error with the Haigis formula was statistically significantly
Koch adjustments reduced the percentage of hyperopic outcomes lower than with Hoffer Q and Holladay 2. The authors report that all
in eyes with axial myopia.19 formulas tend to underestimate the power of the IOL.14
Cheng et  al. evaluated whether the axial length adjustment Hoffer et al. they conducted an evaluation of the Holladay 2
with modified Wang-Koch improved the accuracy of the SRK/T formula’s performance and compared it with the Holladay 1, Hoffer
and Holladay 1 formulas compared to the classic method without Q, and SRK/T formulas in a study involving 317 eyes, which were
correction in eyes with high myopia.11 divided into four categories according to their axial length. Holladay
The study also compared the certainty of the Wang-Koch 2 showed good results in high myopic eyes, while SRK/T consistently
adjustment with the Barrett Universal II formula and the Haigis achieved a lower predictive error in all eyes.14
formula, both of which have been recommended for IOL calculation Yokoi et al., using the SRK/T formula, assessed the refractive
in high myopic eyes in previous studies.11 Barrett Universal II results in patients with an axial length greater than 26.5 mm and
achieved the lowest mean absolute error (MedAE), followed by measured the predictive error. In conclusion, the study found
Haigis with optimized constants and Holladay 1 with modified that a higher incidence of patients with high myopia underwent
Wang-Koch adjustment. 11 phacoemulsification surgery among women and individuals aged
The study’s findings indicated that the modified Wang-Koch approximately 40-50 years. Approximately 80% of patients with
adjustment might be a more favorable choice than optimization high myopia chose mild to moderate myopia as their desired target
for Holladay 1 in larger eyes. However, this was not the case for refraction, while 15% opted for emmetropia.17
the SRK/T formula. Barrett Universal II had the lowest prediction The study conducted by Abulafia et al. It involved a retrospective
error, although Holladay 1 with modified Wang-Koch adjustment review of patients undergoing cataract surgery with an axial length
resulted in a lower percentage of hyperopic outcomes and was greater than 26.00 mm. They analyzed the performance of the
more accurate than the online formula for eyes with axial lengths Holladay 1, Haigis, SRK/T, and Hoffer Q formulas using optical
between 25.00 and 27.00 mm.11 biometry constants, ULIB constants, and axial length adjustment
Fuest et al. In an analysis of refractive outcomes in 63 eyes with methods. They also compared the Holladay 2, Olsen, and Barrett
high myopia that underwent phacoemulsification with intraocular Universal II formulas.16
lens implantation, the study revealed that the precision, measured With the Barrett Universal II formula, a statistically significantly
by MedAE (Mean Absolute Error), was high and comparable for lower numerical error was obtained compared to the standard
Barrett Universal II, Haigis, and RBF formulas, while third-generation formulas.16 The SRK/T, Haigis, Barrett Universal II, Holladay 2, and
formulas Holladay and SRK/T had less accurate performance. Olsen formulas achieved good refractive prediction in axial lengths
Despite this, these formulas remain popular in clinical practice due over 26.00 mm and IOL powers of 6.0D or greater. There was no
to their ability to provide relatively good results with the simplicity benefit in using ULIB constants over standard constants for the
of requiring only a few biometric data points such as keratometry Holladay 1, SRK/T, and Hoffer Q formulas.16
and axial length.15 Several studies have evaluated a substantial number of eyes across
In the study conducted by Idrobo et  al. they compared the different axial length ranges. Aristodemou et al., in their study of 8,108
accuracy of the T2 formula with SRK/T and Holladay 1. They eyes, showed that the SRK/T formula was the most accurate formula
concluded that calculating IOL power in eyes with high myopia for eyes with an axial length greater than 26.00 mm.20
is a challenging task, and even with modern formulas, errors still Kane et al., in their study of 3,241 eyes, showed that the Barrett
exist. SRK/T is one of the most accurate formulas for patients with Universal II formula had a significantly lower absolute error value
high axial length, with the advantage of its availability. Therefore, than all other formulas (Haigis, Hoffer Q, Holladay 1, SRK/T, T2, and
seeking improvement in this method is a relevant issue, even in the Holladay 2).20
presence of new generation formulas. In their study, the T2 formula For third-generation formulas, Holladay 1 had a lower absolute
was less accurate than SRK/T in eyes with high myopia, and they error than Hoffer Q and SRK/T. SRK/T was more accurate than Hoffer
propose a method to improve the performance of the T2 formula.12 Q. Barrett showed significantly higher prediction percentages and
Roessler et  al. in their retrospective study, the researchers lower predictive error than the other formulas in the 77 eyes with
aimed to determine the refractive outcome of cataract surgery and axial length greater than 26 mm, followed by SRK/T, T2, Haigis,
intraocular lens (IOL) implantation in eyes with high axial length. The Holladay 2, Holladay 1, and Hoffer Q.20
study utilized optical biometry and employed the Haigis formula for Zhang et  al. They reported on the accuracy of intraocular
IOL power calculation. They included 37 patients with axial length lens calculation in eyes with axial length greater than 26 mm and
greater than 26.5 mm. Additionally, IOL power predictions were concluded that the Barrett Universal II, SRK/T, and Haigis formulas
recalculated using Holladay Q and SRK/T formulas and compared had similar accuracy; however, the Barrett Universal II formula had
to the results generated by the Haigis formula.18 The results showed a lower predictive error.4
that the Haigis formula had a lower prediction error than Holladay 1, The summary of the selected articles is shown in the following
followed by SRK/T18.18 table (Table 2):

44 Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

 Table 2:  Summary of selected items
AU- OBJETIVO DEL TIPO DE COMPLICA-
TITULO TORES REVISTA AÑO ESTUDIO ESTUDIO MUESTRA DESCRIPCION BREVE LIO CALCULO FORMULA AV RESULTADOS CIONES
1 Comparison of Liu et al. J Cataract 2019 ASESORAR LA SERIE DE 136 OJOS EVALUO FORMULA SN60WF, Al- IOLMaster BARRETT II, NR BARRETT MENOR ERROR, SIN COMPLICA-
intraocular lens Refract EXACTITUD CASOS DE 92 PA- BARRETT UNIVERSAL con Labora- 500, Carl HILL-RBF, SEGUIDO DE HILL-RBF, PERO CIONES TRANS-
power calcula- Surg DE FORMULAS PROSPECTIVO, CIENTES II Y HILL-RBF, HAIGIS, tories, Inc. Zeiss Med- HAIGIS, NO DIFERENCIA SIGNIFICATIVA QUIRURGICAS
tion formulas PARA CAL- ABRIL 2015 A HOLLADAY 1 Y SRK/T. O MA60MA, itec AG HOLLADAY ENTRE BARRETT, HILL-RBF,
in Chinese CULO DE LENTE MARZO 2018 SE AGREGO AJUSTE Alcon Labo- 1, SRK/T HAIGIS Y AJUSTE WANG KOCH
eyes with axial INTRAOCULAR WANG-KOCH ORIGINAL ratories, Inc. CON HOLLADAY 1. FORMULAS
myopia EN OJOS DE PA- Y MODIFICADA CON CON AJUSTE WANG KOCH
CIENTES CHINOS HOLLADAY Q Y SRK/T. MENOR PORCENTAJE DE OJOS
CON LONGITUD SE CALCULO EL ERROR CON RESULTADOS HIPERME-
AXIAL MAYOR A PREDICTIVO RESTANDO TROPES (15-18%) QUE OTRAS
26MM. LA REFRACCION PREDI- FORMULAS (28-91%)
CHA DE LA REFRACCION
POSOPERATORIA.
2 Accuracy Cheng Curren 2019 EVALUAR LA SERIE DE 325 OJOS SE REALIZO FACOEMUL- Akreos Adapt IOL master HOLLADAY NR EL AJUSTE MODIFICADO SIN REPORTE
of Modified et al. Eye Re- EXACTITUD CASOS PRO- DE 238 SIFICACION DE CATARATA AO [Bausch 500, Carl 1, SRK/T Y DE WANG-KOCH OBTUVO DE COMPLICA-
Axial Length search DEL AJUSTE SPECTIVO. DE PACIENTES E IMPLANTE DE LIO EN & Lomb], Tec- Zeiss Med- HAIGIS CON MEJOR DESEMPEÑO QUE LA CIONES
Adjustment MODIFICADO AGOSTO 2016 BOLSA CAPSULAR. OJOS nis ZA9003, itec, Inc. CON- OPTIMIZACION DE CONSTANTE
for Intraocular WANG-KOCH A ABRIL 2018 CON LONGITUD AXIAL Sensar AR40 STANTES PARA HOLLADAY 1 PERO NO
Lens Power PARA CAL- MAYOR A 25 MM USAN- e/E [both Ab- ULIB; PARA SRK/T. HOLLADAY 1 CON
Calculation in CULO DE LENTE DO SRK/T Y HOLLADAY 1 bott Medical HOLLADAY AJUSTE WANG-KOCH FUE
Chinese Axial INTRAOCULAR CON AJUSTE MODIFICA- Optics, Inc.] 1, SRK/T Y MAS CERTERA QUE BARRETT
Myopic Eyes EN OJOS CON DO DE WANG-KOCH Y SE O AcrySof HAIGIS CON PARA LONGITUDES AXIALES
MIOPIA AXIAL COMPARO CON BARRETT SN60WF CON- ENTRE 25.00 Y 27.00 MM. EN
EN EL CENTRO UNIVERSAL II Y HAIGIS. SE [Alcon Labo- STANTES CONCLUSION LOS RESULTA-
OFTALMOLOG- CALCULO ERROR ABSO- ratories, Inc. OPTIMIZA- DOS MOSTRARON QUE EL
ICO, GUANG- LUTO: DIFERENCIA ENTRE DAS; HOL- AJUSTE CON WANG KOCH
ZHOU, CHINA. REFRACCION PREDICHA Y LADAY 1 Y MODIFICADO PUEDE SER
REFRACCION POSOPERA- SRK/T CON UNA MEJOR OPCION QUE LA
TORIA. AJUSTE OPTIMIZACION DE CONSTANTE
MODIFICA- PARA HOLLADAY 1 EN OJOS
DO WANG- LARGOS, PERO NO PARA SRKT.
KOCH Y HOLLADAY 1 + WK MODIFI-
BARRETT II. CADA TUVO MEJOR PORCEN-
TAJE DE HIPERMETROPIA QUE
TODAS LAS DEMAS FORMULAS
INCLUYENDO BARRETT. BAR-
RETT MAYOR PORCENTAJE DE
OJOS DENTRO DE +-0.5D DE
PREDICCION. NO DIFERENCIA
SIGNIFICATIVA ENTRE BARRETT
Y HOLLADAY 1 + WK MOD.
 Contd…

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

45
 Contd…

46
AU- OBJETIVO DEL TIPO DE COMPLICA-
TITULO TORES REVISTA AÑO ESTUDIO ESTUDIO MUESTRA DESCRIPCION BREVE LIO CALCULO FORMULA AV RESULTADOS CIONES
3 Intraocular Fuest Int Oph- Feb- EVALUAR LA SERIE DE 58 ojos SE INCLUYERON 50 OJOS CT Spheris IOL master HAIGIS, 0.29 ± RESULTADOS DE REFRACCION SIN COMPLICA-
lens power et al. thalmol 21 PRECISION DEL CASOS RET- MIOPES CON LONGITUD 204 (Carl 500, Carl BARRETT 0.29 DENTRO DE 1 DIOPTRIA DE CIONES TRANS-
calculation CALCULO DE ROSPECTIVO AXIAL >26.00MM, A Zeiss Med- Zeiss Med- UNIVERSAL log- REFRACCION META: BARRETT OPERATORIAS.
for plus and LA POTENCIA QUIENES SE LES REALIZÓ itec) itec, Inc. II, HOL- MAR 80%, HAIGIS 87%, RBF 82%.
minus lenses DE LENTE FACOEMULSIFICACION LADAY 1, LA EXACTITUD DE BARRETT,
in high myopia INTRAOCULAR E IMPLANTE DE LENTE HILL-RBF Y HAIGIS Y RBF FUERON COMPA-
using partial EN PACIENTES INTRAOCULAR. TRAS SRK/T RABLES CON UNA TENDENCIA
coherence CON MIOPIA BIOMETRIA USANDO A MAYOR ERROR ABSOLUTO
interferometry ALTA, IMPLAN- IOLMASTER 500. SE EN LENTES INTRAOCULARES
TANDO LENTES REALIZO EL CALCULO NEGATIVOS.
POSITIVAS Y CON FORMULA DE
NEGATIVAS. HAIGIS,PARA COMPARA-
CION LA REFRACCION
SE CALCULO USANDO
BARRETT UNIVERSAL II,
HOLLADAY 1, HILL-RBF
Y SRK/T. SE UTILIZARON
CONSTANTES OPTIMIZA-
DAS ULIB.
4 T2 formula in a Idrobo- BMC 2019 DETERMINAR SERIE DE 63 ojos 63 OJOS MIOPES ALTOS A Alcon IOL master T2, SRK/T, NR FORMULA T2 PARECE SER NO SE REPOR-
highly myopic Robali- Ophthal- EXACTITUD DE CASOS RET- QUIENES SE LES REALIZO Acrisoft® 500, Carl HOLLADAY MENOS EXACTA QUE LA TAN LAS COM-
population, no et al. mology FORMULA T2 ROSPECTIVO, FACOEMULSIFICA- SN60WF Zeiss Med- 1. FORMULA SRK/T EN OJOS CON PLICACIONES
comparison APLICADA EN JUN 2012 A CION DE CRISTALINO E itec, Inc. ALTA MIOPIA.
with other ALTOS MIOPES, NOV 2015 INSERCION DE LENTE
methods and COMPARAR INTRAOCULAR. SE OBTU-
description of FORMULA T2 VIERON LOS ERRORES DE
an improved CON SRK/T Y PREDICCION.
approach for HOLLADAY 1 Y
estimating DESCRIBIRPOSI-
corneal height BLES FIRMAS

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

DE MEJORAR
LA ESTIMA-
CION ALTURA
CORNEAL Y
PREDICCION DEL
ERROR.
5 Accuracy of Roessler Ophthal- Feb- ASESORAR LA SERIE DE 37 OJOS SE REALIZO FACO- Acrysof MA- Zeiss HAIGIS, NR EL ERROR EN PREDICCION FUE NO SE REPOR-
intraocular lens et al. mic & 12 EXACTITUD CASOS RET- EMULSIFICACION CON 60BM (Alcon IOLMaster SRK/T, HOL- SIGNIFICATIVAMENTE MENOR TAN
power calcula- Physi- DEL CALCULO ROSPECTIVO, IMPLANTE DE LENTE Laboratories) (Version LADAY 1 CON FORMULA DE HAIGIS EN
tion using par- ological DE LENTE IN- INTRAOCULAR EN OJOS Y ACR6D 3.01; Carl UN MAYOR NUMERO DE SUJE-
tial coherence Optics TRAOCULAR MIOPES CON LONGITUD SE (Corneal Zeiss TOS DENTRO DE LOS LIMITES
interferometry CON BIOMETRIA AXIAL MAYOR A 26.5 MM, Laboratories) Meditec, DE REFRACCION COMPARADO
in patients with OPTICA USANDO UTILIZANDO IOLMASTER CON LAS DEMAS FORMUNLAS.
high myopia IOL MASTER EN PARA EL CALCULO DE
UN GRUPO DE LENTE INTRAOCULAR. SE
PACIENTES CON UTILIZO FORMULA DE
MIOPIA. HAIGIS EN TODOS LOS
CASOS. LA REFRACCION
FUE RE-CALCULADA
USANDO LAS FORMULAS
SRK/T Y HOLLADAY 1 Y
FUE COMPARADA CON
LOS RESULTADOS INI-
CIALES GENERADOS POR
IOLMASTER USANDO
Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

FORMULA HAIGIS.
 Contd…
 Contd…
AU- OBJETIVO DEL TIPO DE COMPLICA-
TITULO TORES REVISTA AÑO ESTUDIO ESTUDIO MUESTRA DESCRIPCION BREVE LIO CALCULO FORMULA AV RESULTADOS CIONES
6 Accuracy Terzi J Cataract 2009 DETERMINAR LA SERIE DE 44 OJOS SE REALIZO FACOEMULSI- AcrySof IOL master Hol- NR CON LAS CONSTANTES OPTIMI- NO SE REPOR-
of modern et al. Refract EXACTITUD DE CASOS RET- MIOPES FICACION CON IMPLANTE MA60MA 500, Carl laday IOL ZADAS, LA FORMULA HAIGIS TAN
intraocular lens Surg LAS FORMULAS ROSPECTIVO, DE 27 PA- DE LENTE INTRAOCULAR IOL, AcrySof Zeiss Med- Consultant PRODUJO EL ERROR PREDIC-
power calcula- HOLLADAY NOVIEMBRE CIENTES EN CAMARA POSTE- MA60BM, itec, Inc. software TIVO MAS BAJO, SEGUIDO DE
tion formulas 2, HOFFER Q, 200 A DICIEM- RIOR EN 44 OJOS CON AcrySof (version SRK/T, HOLLADAY 2 Y HOFFER
in refractive SRK/T Y HAIGIS BRE 2004 MIOPIA ALTA Y 19 CON SA60AT, 2.50.1339). Q. EL ERROR CON FORMULA
lens exchange PARA CALCULO HIPERMETROPIA ALTA. EL Acry- Sof Y PARA DE HAIGIS FUE ESTADISTICA-
for high myo- DE LENTE IN- ERROR DE PREDICCION MA60AC, COMPARA- MENTE SIGNIFICATIVO MAS
pia and high TRAOCULAR EN SE CALCULO DE FORMA Sensar SION DE BAJO QUE CON HOFFER Q Y
hyperopia INTERCAMBIO RETROSCPECTIVA PARA AR40e FORMA HOLLADAY 2.
REFRACTIVO DE LA FORMULA HOLLADAY RETRO-
CRISTALINO EN 2, HOFFER Q, SRK/T Y SPECTIVA
MIOPIA ALTA E HAIGIS. HOLLADAY
HIPERMETRO- 2, HOFFER
PIA ALTA CON Q, SRK/T Y
LIO FLEXIBLE HAIGIS.
DE ACRILICO
HIDROFOBICO.
7 Evaluation of Yokoi Int Oph- 2013 ANALIZAR ER- SERIE DE 84 OJOS FACOEMULSIFICACION AcrySof 3 IOL Master SRK/T NR EL ERROR REFRACTIVO POSOP- NO SE REPOR-
refractive error et al. thalmol ROR REFRAC- CASOS RET- DE 64 PA- DE CRISTALINO CON piezas (software ERATORIO FUE SIGNIFICATIVA- TAN
after cataract TIVO META Y ROSPECTIVO, CIENTES IMPLANTE DE LENTE version MENTE MAYOR EN QUELLOS
surgery in POSOPERATORIO ABRIL 2009 A INTRAOCULAR EN 84 3.01.0294; OJOS CON LONGITUD AXIAL
highly myopic EN OJOS CON MARZO 2010 OJOS CON MIOPIA ALTA Carl Zeiss >31.00D.
eyes MIOPIA ALTA (MAYOR A 26.50MM). Meditec
TRAS FACOE- SE REALIZO EXPLORA- AG, Jena,
MULSIFICACION CION OFTALMOLOGICA Germany)
E IMPLANTE COMPLETA INCLUYENDO
DE LENTE AV MEJOR CORREGIDA,
INTRAOCULAR Y ERROR REFRACTIVO
EXAMINAR LOS (EQUIVALENTE ESFERICO)
FACTORES PRE- Y OFTALMOSCOPIA POR
OPERATORIOS LO MENOS UNA VEZ
QUE INFLUEN- CADA 3 MESES POR UN
CIARAN ESTOS AÑO DE SEGUIMIENTO.
DEFECTOS
REFRACTIVOS.
 Contd…

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

47
 Contd…

48
AU- OBJETIVO DEL TIPO DE COMPLICA-
TITULO TORES REVISTA AÑO ESTUDIO ESTUDIO MUESTRA DESCRIPCION BREVE LIO CALCULO FORMULA AV RESULTADOS CIONES
8 Intraocular lens Abulafia J Cataract 2015 EVALUAR Y SERIE DE 106 OJOS SE REALIZO FACO- AcrySof IOLMaster, HOLLADAY NR LAS FORMULAS SRK/T, HAIGIS NO SE REPOR-
power calcula- et al. Refract COMPARAR LA CASOS RET- de 68 EMULSIFICACION CON (Alcon) software 1, SRK/T, (ULIB), BARRETT UNIVERSALL II, TAN
tion for eyes Surg EXACTITUD DE ROSPECTIVO. pacientes IMPLANTE DE LENTE V5.4 y HOFFER HOLLADAY 2 Y OLSEN TIENEN
with an axial LAS FORMULAS OCT 2010 A INTRAOCULAR, SE COM- Lenstar Q, HAIGIS, LA MEJOR PREDICCION DE
length greater Y ETODOS PARA JULIO 2012 PRARO EL RESULTADO biometry BARRET RESULTADOS REFRACTIVOS
than 26.0 mm: CALCULO DE REFRACTIVO POSOPERA- (Haag- UNIVERSAL PARA LONGITUDES AXIALES
Comparison of LENTE INTRACU- TORIO CON LOS VALORES Streit AG) II, HOL- MAYORES A 26.00 MM Y PARA
formulas and LAR PARA OJOS PREDICHOS USANDO LAS LADAY 2, PODER DE LENTE INTRAOCU-
methods CON LONGITUD FORMULAS: HOLLADAY OLSEN LAR DE 6.00 D O MAYOR. NO SE
AXIAL MAYOR A 1, SRK/T, HOFFER Q Y HAI- VIO BENEFICIO EN USAR LAS
26 MM. GIS CON CONSTANTES CONSTANTES ULIB A COMPAR-
IPTICAS DEL LENTE, CON ACION DE LAS CONSTANTES
CONSTANTES DEL USER ESTANDAR PARA HOLLADAY 1,
GROUP FOR LASER IN- SRK/T Y HOFFER Q. LA MEJOR
TERFERENCE BIOMETRY PREDICCION SE OBTUVO CON
CONSTANTS Y CON EL HOLLADAY 1 ( CON AJUSTE DE
METODO DE AJUSTE LA), HAIGIS ( CON AJUSTE DE
DE LONGITUD AXIAL LA) Y BARRET UNIVERSAL II.
USANDO FORMULAS
DE NUEVA GENERA-
CION COMO BARRETT
UNIVERSAL II, HOLLADAY
2 Y OLSEN.
9 Intraocular lens Kane J Cataract 2016 ASESORAR LA SERIE DE 3241, 77 DATOS DE PACIENTES Acrysof IQ IOLMaster HAIGIS, NR BARRETT II OBTUVOEL MAE NO SE REPOR-
power formula et al. Refract EXACTITUD DE 7 CASOS RET- CON MIO- A QUIENES SE REALIZO SN60WF (version HOFFER Q, MAS BAJO DE LAS FORMULAS TAN
accuracy: Com- Surg FORMULAS UTI- ROSPECTIVO, PIA ALTA CIRUGIA DE CATARATA 5.4, Carl HOLLADAY EVALUADAS, LA DIFERENCIA
parison of 7 LIZADAS PARA FEB 2010 A SIN EVENTUALIDADES Zeiss Med- 1, SRK/T, FUE ESTADISTICAMENTE SIG-
formulas CALCULAR EL NOV 2015 Y SE IMPLANTO LENTE itec AG) BARRETT NIFICATIVA COMPARADA CON
PODER DE LENTE INTRAOCULAR, SE UNIVERSAL HAIGIS, HOFFER Q, HOLLADAY

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

INTRAOCU- USARON CONSTANTES II, T2, HOL- Q, HOLLADAY 2 Y SRK/T.
LAR USANDO OPTIMIZADAS PARA LADAY 2
BIOMETRIA POR DETERMINAR EL ERROR
IOLMASTER Y EN LA PREDICCION DEL
CONSTANTES DE RESULTADO REFRACTIVO.
LENTES OPTIMI- OJOS SEPARADOS EN
ZADAS. SUBGRUPOS SEGÚN
LONGITUD AXIAL: CORTO
(<22MM), MEDIANO
(>22-<24.5), MEDIANO-
LARGO (>24.5 A <26.00) Y
LARGOS (>26.00MM)
10 Effect of Axial Zhang Am J 2020 EVALUAR EL SERIE DE 164 OJOS PACIENTE CON MIOPIA enVista IOLMaster HOLLADAY NR LAS FORMULAS HOLLADAY 1 NO SE REPOR-
Length Adjust- et al. Ophthal- DESEMPEÑO DE CASOS RET- DE 164 ALTA Y LONGITUD model MX60; 700 (Carl 1, SRK/T Y SRK/T COMBINADAS CON TAN
ment Methods mol LAS FORMULAS ROSPECTIVO. PACIENTES AXIAL MAYOR A 26 MM A Bausch & Zeiss METODOS DE AJUSTE DE LON-
on Intraocular HOLLADAY 1 Y QUIENES SE LES REALIZO Lomb Incor- Medi- tec, GITUD AXIAL CONTARON CON
Lens Power SRK/T JUNTO CIRUGIA POR FACOEMUL- porated Jena, Ger- PRESICION SOMLAR A FORMU-
Calculation in CON LOS METO- SIFICACION E IMPLANTE many) LAS DE CUARTA GENERACION
Highly Myopic DOS DE AJUSTE DE LENTE INTRAOCULAR. EN OJOS CON MIOPIA ALTA.
Eyes DE LONGITUD
AXIAL EN
PACIENTES CON
MIOPIA ALTA
Y LONGITUD
Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

AXIAL MAYOR A
26.00 MM.
 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

The selected articles that assessed the accuracy of postoperative M e ta - a n a lys i s R e s u lts
refractive prediction by calculating the prediction error using MAE
(Mean Absolute Error) are presented below (Table 3). Single-Mean Meta-Analysis for MAE
The data from the most frequently found formulas in the For this estimator, seven studies were included that reported the
selected studies, including MAE (Mean Absolute Error), MNE Mean Absolute Error (MAE) (Terzi 2011, Ressler 2012, Ressler 2013,
(Mean Numerical Error), or MedAE (Median Absolute Error), are Yokoi 2013, Abulafia 2015, Idrobo_Robalino 2019, and Zhang 2020)
as follows (Table 4). (Table 5).

Tabla 3:   Predictive results of selected articles

ITEM HAIGIS BARRETT SRK/T HOFFER Q

N MAE RANGE MAE RANGE MAE RANGE MAE RANGE

1 136 0.41 0.32 0.49
2 325 0.61 0.48
3 58 0.38 0.25-0.63 0.49 0.34-0.64 0.44 0.25-0.75
4 63 0.42
5 37 0.7 1.01
6 44 0.21 0.01-0.91 0.23 0.00-0.73 0.32 0.04-0.91
7 84 0.72 0.25-1.19
8 76 0.31 0.00-0.80 0.28 0.00-0.71 0.28 0.01-0.98 0.37 0.02-1.36
8 30 0.69 0.06-1.79 0.3 0.03-1.18 0.84 0.20-2.10 1.42 0.30-2.57
9 77 0.526 0.435 0.48 0.589
10 164 0.38 0.43
ITEM HOLLADAY 1 HOLLADAY 2 OLSEN HILL -RBF T2
MAE RANGE MAE RANGE MAE RANGE MAE RANGE MAE RANGE
1 0.45 0.37
2 0.73
3 0.75 0.40-1.13 0.44 0.25-0.57
4 0.455 0.435
5 0.85
6 0.29 0.01-0.92
7
8 0.4 0.00-1.55 0.34 0.01-1.24 0.26 0.01-0.81
8 1.21 0.23-2.29 1.13 0.07-2.33 0.49 0.05-1.37
9 0.586 0.544 0.498
10 0.62
ITEM HOLLADAY 1 /WANG-KOCH SRK/T /WANG-KOCH HOLLADAY 1 /WANG-KOCH MOD SRK/T /WANG-KOCH MOD
MAE RANGE MAE RANGE MAE RANGE MAE RANGE
1 0.37 0.46 0.39 0.47
2 0.47 0.55
3
4
5
6
7
8
9
10

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023) 49

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

Table 4:  Results of MAE, MNE and MedAE

ITEM HAIGIS BARRETT
N MAE DS RANGO MNE DS MedAE MAE DS RANGO MNE DS MedAE
1 136 0.41 0.29 0.52 0.36 0.32 -0.09 0.42 0.27
2 325 0.61 0.00 0.84 0.42 0.48 -0-08 0.67 0.34
3 58 0.38 0.49
4 63
5 37 0.7 0.59
6 44 0.21 0.21 0.01-0.91 -0.75 0.52
7 84
8 76 0.31 0.22 0.00-0.80 -0.17 0.35 0.3 0.28 0.19 0.00-0.71 -0.1 0.32 0.26
8 30 0.69 0.38 0.06-1.79 0.67 0.41 0.66 0.3 0.21 0.03-1.18 0.1 0.39 0.21
9 77 0.526 0.392 0.435 0.37
10 164 0.38 0.49 0.28

ITEM SRK/T HOLLADAY 1

N MAE DS RANGO MNE DS MedAE MAE DS RANGO MNE DS MedAE
1 136 0.49 0.25 0.68 0.34 0.45 0.7 0.56 0.4
2 325 0.73 0 0.93 0.58
3 58 0.44 0.75
4 63 0.418 0.327 0.352 0.455 0.314 0.389
5 37 1.01 0.61 0.85 0.68
6 44 0.23 0.18 0.00-0.73 -0.33 0.44
7 84 0.72 0.47 0.25-1.19
8 76 0.28 0.21 0.01-0.98 -0-05 0.35 0.23 0.4 0.31 0.00-1.55 0.35 0.36 0.35
8 30 0.84 0.5 0.20-2.10 0.82 0.53 0.67 1.21 0.41 0.23-2.29 1.21 0.41 1.19
9 77 0.484 0.419 0.586 0.441
10 164 0.43 0.58 0.35 0.62 0.48 0.4

ITEM HOLLADAY 2
N MAE DS RANGO MNE DS MedAE
1 136
2 325
3 58
4 63
5 37
6 44 0.29 0.2 0.01-0.92 -0.5 0.47
7 84
8 76 0.34 0.28 0.01-1.24 0.22 0.38 0.29
8 30 1.13 0.47 0.07-2.33 1.13 0.47 1.02
9 77 0.544 0.404
10 164

50 Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

Fig. 2:  Meta-analysis results of single means for MAE

Barrett’s Universal Formula II mean (RM) across the studies. The overall MAE estimate with the
Two studies used the Barrett Universal II formula (Abulafia Haigis formula was 0.37 (95% CI = 0.21; 0.53).
2015 and Zhang 2020). The estimator of variance (Tau2) among
SRK-T Formula
these combined studies showed significant heterogeneity
(Tau2 = 0.004, Chi2 = 5.16, df = 1, p = 0.02), and I2 indicated Five studies used the SRK-T formula (Terzi 2011, Ressler 2013, Yokoi
81% heterogeneity). The random-effects model was used to 2013, Abulafia 2015, Idrobo_Robalino 2019). The estimator of
estimate the overall pooled mean (RM) across the studies. The variance (Tau2) among these combined studies showed significant
overall MAE estimate with the Barrett formula was 0.33 (95% heterogeneity (Tau2 = 0.0424, Chi2 = 125.76, df = 4, p <0.01), and
CI = 0.23; 0.42). I2 indicated 97% heterogeneity. The random-effects model was
used to estimate the overall pooled mean (RM) across the studies.
The overall MNE estimate with the SRK-T formula was 0.51 (95%
CI = 0.33; 0.70) (Fig. 2).
Haigis Formula
Three studies used the Haigis formula (Ressler 2012, Terzi 2009,
and Abulafia 2015). The estimator of variance (Tau2) among these Single -Mean Meta-Analysis for MNE
combined studies showed significant heterogeneity (Tau2 = 0.0166, For this estimator, six studies were included that reported the Mean
Chi2 = 24.87, df = 2, p <0.01), and I2 indicated 92% heterogeneity. Numerical Error (MNE) (Terzi 2010, Terzi 2011, Abulafia 2015, Liu 2019,
The random-effects model was used to estimate the overall pooled Cheng 2019, and Liu 2020) (Fig. 3).

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023) 51

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

Fig. 3:  Metanalysis results for single means for MNE

Barrett’s Universal Formula II SRK-T Formula

Three studies used the Barrett Universal II formula (Abulafia Three studies used the SRK-T formula (Terzi 2011, Abulafia 2015,
2015, Cheng 2019, and Liu 2019). The estimator of variance Liu 2020). The estimator of variance (Tau2) among these combined
(Tau2) among these combined studies did not show significant studies showed significant heterogeneity (Tau2 = 0.0611, Chi2 =
heterogeneity (Tau2 = 0, Chi2 = 0.15, df = 2, p = 0.93), and I2 43.76, df = 2, p <0.01), and I2 indicated 95% heterogeneity. The
indicated 0% heterogeneity. The fixed-effects model was used random-effects model was used to estimate the overall pooled
to estimate the overall pooled mean (RM) across the studies. mean (RM) across the studies. The overall MAE estimate with the
The overall MNE estimate with the Barrett formula was -0.09 SRK-T formula was -0.04 (95% CI = -0.33; 0.24) (Fig. 3).
(95% CI = -0.13; -0.05).
Haigis Formula Discussion
Four studies used the Haigis formula (Terzi 2010, Abulafia 2015, Liu The accurate calculation of intraocular lens power is a debated topic,
2019, Cheng 2019). The estimator of variance (Tau2) among these with no previous consensus on which formula is the most accurate
combined studies showed significant heterogeneity (Tau2 = 0.1664, for calculating in patients with high myopia.
Chi2 = 147.95, df = 3, p <0.01), and I2 indicated 98% heterogeneity. The Barrett Universal II is a fourth-generation formula capable
The random-effects model was used to estimate the overall pooled of predicting the posterior corneal curvature. It is highly effective
mean (RM) across the studies. The overall MNE estimate with the in different ranges of axial lengths, especially in high myopia. It
Haigis formula was -0.15 (95% CI = -0.49; 0.19). uses a lens factor that considers both the physical position and the

52 Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

location of the principal planes of the intraocular lens, which is an studies related to new formulas or modifications applied to
advantage over other formulas used for intraocular lens calculation. patients with high myopia, it is suggested to compare them with
It obtained the lowest mean absolute prediction error compared the Barrett Universal II formula, as it provides the most predictable
to other similar formulas, as assessed by two studies (Abulafia 2015 and accurate results in calculating intraocular lens power in this
and Zhang 2020) included in the meta-analysis of single means for patient group.
MAE. This formula achieved a global MAE estimate of 0.33 (0.23; 0.42),
being the formula that obtained the lowest global mean absolute D e c l a r at i o n of Conflict of Interest
error estimator, compared to the Haigis formula with a global MAE
The authors have declared no potential conflicts of interest in regard
of 0.37 (0.21; 0.53) and SRK/T with 0.51 (0.33; 0.70).
to the research, authorship, and/or publication of this article.
In the meta-analysis of single means for MNE, the results were
consistent with those obtained by MAE measurement, with the
Barrett Universal II formula obtaining a lower overall value of mean
Sponsors
numerical error. The values obtained were -0.09 (-0.13; -0.05) for the The authors received no financial support for the research,
Barrett formula, -0.15 (-0.49; 0.19) for the Haigis formula, and -0.04 authorship, and/or publication of this article.
(-0.33; 0.24) for SRK/T. Therefore, Barrett Universal II is the formula
with the least error in calculating intraocular lens power in patients References
with high myopia, regardless of the estimator used (MNE or MAE). 1. Harb EN, Wildsoet CF. Origins of Refractive Errors: Environmental
The use of the Holladay 1 and SRK/T formulas combined with and Genetic Factors. Annu Rev Vis Sci. 2019;5(1):47–72. DOI: 10.1146/
specific adjustments for axial length has a similar accuracy to fourth- annurev-vision-091718-015027
2. Holden BA, Fricke TR, Wilson DA, et  al. Global Prevalence of
generation formulas in patients with high myopia, as reported by
Myopia and High Myopia and Temporal Trends from 2000 through
Zhang et al.4 Therefore, the advantage of the Barrett Universal II 2050. Ophthalmology. 2016;123(5):1036–1042. DOI: 10.1016/j.
formula is that it does not require modifications for its application ophtha.2016.01.006
in this group of patients. 3. Wu PC, Huang HM, Yu HJ, Fang PC, Chen CT. Epidemiology of
Although the Haigis formula provides satisfactory results in Myopia: Asia-Pac J Ophthalmol. 2016;5(6):386–393. DOI: 10.1097/
calculating power for myopic patients,18 further research on a larger APO.0000000000000236
number of eyes with high myopia requiring negative power lenses 4. Zhang J, Tan X, Wang W, et  al. Effect of Axial Length Adjustment
may be beneficial to identify the reasons for the lack of accuracy Methods on Intraocular Lens Power Calculation in Highly Myopic Eyes.
Am J Ophthalmol. 2020;214:110–118. DOI: 10.1016/j.ajo.2020.02.023
in these special cases.
5. Gabrić N, Dekaris I, Karaman Ž. Refractive Lens Exchange for
The Haigis formula with optimized constants performed better Correction of High Myopia. Eur J Ophthalmol. 2002;12(5):384–387.
in eyes with axial myopia according to Terzi et al. The results were DOI: 10.1177/112067210201200507
better than previous studies evaluating this formula, which is 6. Lyle AW, Jin GJC. Clear lens extraction for the correction of high
likely attributable to the younger age of the included subjects, refractive error. J Cataract Refract Surg. 1994;20(3):273–276.
optimization of lens constants, and the fact that only patients DOI: 10.1016/S0886-3350(13)80577-9
without cataracts were included in this study. All formulas tended 7. Packard R . Refrac tive lens exchange for myopia: a new
to underestimate the intraocular lens power, which was more p e r s p e c t i ve?: Cu r r O p i n O p ht h a l m o l . 20 05;16 (1): 53 –56 .
DOI: 10.1097/00055735-200502000-00009
severe in eyes with an axial length greater than 31 mm, as many
8. Srinivasan B, Leung HY, Cao H, Liu S, Chen L, Fan AH. Modern
of these eyes had posterior staphyloma detected preoperatively Phacoemulsification and Intraocular Lens Implantation (Refractive
using B-mode ultrasound. Lens Exchange) Is Safe and Effective in Treating High Myopia:
Relatively, a small portion of the population has eyes with Asia-Pac J Ophthalmol. 2016;5(6):438 – 4 4 4. DOI: 10.1097/
an axial length equal to or greater than 26.0 mm. Therefore, in APO.0000000000000241
real-world conditions, most surgeons will not have access to an 9. Fernández-Vega L, Alfonso JF, Villacampa T. Clear lens extraction for
optimized constant for each type of intraocular lens16. the correction of high myopia. Ophthalmology. 2003;110(12):2349–
Based on the results obtained in the meta-analysis, it can be 2354. DOI: 10.1016/S0161-6420(03)00794-2
10. Ramos M, Kruger EF, Lashkari K. Biostatistical analysis of pseudophakic
concluded that the Barrett Universal II formula is the most accurate
and aphakic retinal detachments. :8.
and reproducible for calculating intraocular lens power in patients 11. Cheng H, Liu L, Sun A, Wu M. Accuracy of Modified Axial
with high myopia. It also has high accessibility as it is used in various Length Adjustment for Intraocular Lens Power Calculation in
biometric platforms and is freely available on the web. Chinese Axial Myopic Eyes. Curr Eye Res. 2020;45(7):827–833.
Using the Mean Absolute Error (MAE) estimator to evaluate the DOI: 10.1080/02713683.2019.1698053
prediction accuracy of a formula is recommended since it considers 12. Idrobo-Robalino CA, Santaella G, Gutiérrez ÁM. T2 formula in a
absolute values, eliminating the bias of symbols in the results of the highly myopic population, comparison with other methods and
differences between the actual value and the prediction. description of an improved approach for estimating corneal height.
BMC Ophthalmol. 2019;19(1):222. DOI: 10.1186/s12886-019-1226-7
13. Wang L, Shirayama M, Ma XJ, Kohnen T, Koch DD. Optimizing
C o n c lu s i o n s intraocular lens power calculations in eyes with axial lengths
Calculating intraocular lens power in eyes with high myopia remains above 25.0 mm. J Cataract Refract Surg. 2011;37(11):2018–2027.
a challenging issue, and even with modern formulas, errors persist. DOI: 10.1016/j.jcrs.2011.05.042
14. Terzi E, Wang L, Kohnen T. Accuracy of modern intraocular lens power
This reality highlights the importance of continuous research and
calculation formulas in refractive lens exchange for high myopia
improvements in this matter. and high hyperopia. J Cataract Refract Surg. 2009;35(7):1181–1189.
There were insufficient data for inclusion in the meta-analysis DOI: 10.1016/j.jcrs.2009.02.026
for formulas that incorporated modifications for axial length, 15. Fuest M, Plange N, Kuerten D, et  al. Intraocular lens power
Wang-Koch adjustment, T2, Hill-RBF, and Olsen. If conducting calculation for plus and minus lenses in high myopia using partial

Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023) 53

 Formulas for Intraocular lens Calculation in Phacorefractive Surgery of Patients with high Myopia. Which is the most Accurate?

coherence interferometry. Int Ophthalmol. 2021;41(5):1585–1592. patients with high myopia: IOL power calculation in high myopia.
DOI: 10.1007/s10792-020-01684-y Ophthalmic Physiol Opt. 2012;32(3):228–233. DOI: 10.1111/j.1475-
16. Abulafia A, Barrett GD, Rotenberg M, et al. Intraocular lens power 1313.2012.00903.x
calculation for eyes with an axial length greater than 26.0 mm: 19. Liu J, Wang L, Chai F, et  al. Comparison of intraocular lens power
Comparison of formulas and methods. J Cataract Refract Surg. calculation formulas in Chinese eyes with axial myopia. J Cataract
2015;41(3):548–556. DOI: 10.1016/j.jcrs.2014.06.033 Refract Surg. 2019;45(6):725–731. DOI: 10.1016/j.jcrs.2019.01.018
17. Yokoi T, Moriyama M, Hayashi K, Shimada N, Ohno-Matsui K. 20. Kane JX, Van Heerden A, Atik A, Petsoglou C. Intraocular lens power
Evaluation of refractive error after cataract surgery in highly myopic formula accuracy: Comparison of 7 formulas. J Cataract Refract Surg.
eyes. Int Ophthalmol. 2013;33(4):343–348. DOI: 10.1007/s10792-012- 2016;42(10):1490–1500. DOI: 10.1016/j.jcrs.2016.07.021
9690-6 21. Pierdant M, Castillo A, Tirado R. Cómo leer un artículo científico
18. Roessler GF, Dietlein TS, Plange N, et  al. Accuracy of intraocular ciencias de la salud. UASLP. Vol. 1. San Luis Potosí: UASLP; 1AD.
lens power calculation using partial coherence interferometry in

54 Revista Mexicana de Oftalmología (ENG), Volume 97 Issue 2 (March–April 2023)