Kératomileusis in situ assisté par laser (LASIK) avec un microkératome mécanique comparé au LASIK avec un laser femtoseconde pour le LASIK chez les adultes atteints de myopie ou d'astigmatisme myopique
Résumé scientifique
Contexte
Le LASIK (laser‐assisted in‐situ keratomileusis) est une procédure chirurgicale qui corrige les erreurs de réfraction. Cette technique consiste à créer un volet des parties les plus externes de la cornée (épithélium, membrane de Bowman et stroma antérieur) pour exposer la partie médiane de la cornée (lit stromal) et la remodeler au moyen d'un laser excimer par photoablation. Les volets peuvent être créés par un microkératome mécanique ou un laser femtoseconde.
Objectifs
Comparer l'efficacité et la tolérance du microkératome mécanique par rapport au laser femtoseconde dans le cadre du LASIK pour les adultes myopes.
Stratégie de recherche documentaire
Nous avons fait des recherches sur CENTRAL (qui contient le registre d’essais du groupe Cochrane sur l’ophtalmologie) (2019, numéro 2) ; Ovid MEDLINE ; Embase ; PubMed ; LILACS ; ClinicalTrials.gov et le Système d'enregistrement international des essais cliniques (ICTRP) de l'Organisation mondiale de la santé (OMS). Nous n'avons utilisé aucune restriction de date ou de langue. Nous avons recherché les listes bibliographiques des essais inclus. Nous avons effectué une recherche dans les bases de données électroniques le 22 février 2019.
Critères de sélection
Nous avons inclus des essais contrôlés randomisés (ECR) du LASIK avec un microkératome mécanique comparé à un laser femtoseconde chez des personnes âgées de 18 ans ou plus présentant une myopie ou un astigmatisme myopique de plus de 0,5 dioptrie.
Recueil et analyse des données
Nous avons suivi les procédures méthodologiques standard définies par Cochrane.
Résultats principaux
Nous avons inclus 16 enregistrements de 11 essais enrôlant 943 adultes (1691 yeux) atteints de myopie sphérique ou sphérocylindrique, qui étaient des candidats possibles pour le LASIK. 547 participants (824 yeux) ont reçu le LASIK avec un microkératome mécanique et 588 participants (867 yeux) avec un laser femtoseconde. Chaque essai comprenait entre neuf et 360 participants. Dans six essais, les mêmes participants ont bénéficié des deux interventions. Dans l'ensemble, les essais étaient exposés à un risque incertain de biais pour la plupart des domaines.
À 12 mois, les données d'un essai (42 yeux) n'indiquent aucune différence dans l'acuité visuelle moyenne non corrigée (échelle logMAR) entre le LASIK avec un microkératome mécanique et le LASIK avec un laser femtoseconde (différence moyenne (DM) ‐0,01, intervalle de confiance (IC) à 95% ‐0,06 à 0,04 ; données probantes d’un niveau de confiance faible). Des résultats similaires ont été observés 12 mois après l'opération, concernant les participants ayant atteint 0,5 dioptrie dans la réfraction cible (rapport de risque (RR) 0,97, 95 % IC 0,85 à 1,11 ; 1 essai, 79 yeux ; données probantes d’un niveau de confiance faible) ainsi que l'équivalent sphérique moyen de l'erreur de réfraction 12 mois après l'opération (DM 0,09, 95 % IC ‐0,01 à 0,19 ; 3 essais, 168 yeux [92 participants] ; données probantes d’un niveau de confiance faible).
Sur la base des données de trois essais (134 yeux, 67 participants), le microkératome mécanique a été associé à un risque moindre de kératite lamellaire diffuse par rapport au laser femtoseconde (RR 0,27, 95 % IC 0,10 à 0,78 ; données probantes d’un niveau de confiance faible). Ainsi, la kératite lamellaire diffuse était un événement indésirable plus fréquent avec le laser femtoseconde qu'avec le microkératome mécanique, passant d'un taux supposé de 209 pour 1000 personnes dans le groupe du laser femtoseconde à 56 pour 1000 personnes dans le groupe du microkératome mécanique. Les données d'un essai (183 yeux, 183 participants) indiquent que la sécheresse oculaire en tant qu'événement indésirable pourrait être plus fréquente avec le microkératome mécanique qu'avec le laser femtoseconde, passant d'un taux supposé de 80 pour 1000 personnes dans le groupe du laser femtoseconde à 457 pour 1000 personnes dans le groupe du microkératome mécanique (RR 5,74, 95 % IC 2,92 à 11,29 ; données probantes de faible niveau de certitude). Aucune différence n'a été constatée entre les deux groupes pour le voile cornéen (RR 0,33, 95 % IC 0,01 à 7,96 ; 1 essai, 86 yeux) et la croissance épithéliale (RR 1,04, 95 % IC 0,11 à 9,42 ; 2 essais, 102 yeux [51 participants]). Le niveau de confiance des données probantes pour les deux critères de jugement était très faible.
Conclusions des auteurs
En ce qui concerne les résultats sur l'acuité visuelle, il n'y a peut‐être pas de différence entre le LASIK avec microkératome mécanique et le LASIK avec laser femtoseconde. La sécheresse oculaire et la kératite lamellaire diffuse sont des effets indésirables probables avec le microkératome mécanique et le laser femtoseconde. Les données probantes sont incertaines en ce qui concerne l’opacité cornéenne et la croissance épithéliale interne comme événements indésirables de chaque intervention. Le nombre limité de critères de jugement rapportés dans les essais inclus, dont certains présentant un risque de biais potentiellement important, rend difficile de tirer une conclusion ferme concernant l'efficacité et la tolérance des interventions étudiées dans le cadre de cette revue.
PICOs
Résumé simplifié
Efficacité et sécurité de deux types d'outils utilisés dans le cadre du LASIK (un type de chirurgie réfractive) pour la myopie
Quel est l’objectif de cette revue ?
L'objectif de cette revue Cochrane était d'évaluer si deux types d'outils de correction chirurgicale de la vision (LASIK ‐ laser‐assisted in‐situ keratomileusis) sont efficaces et sûrs chez les personnes myopes. Un outil utilise une lame de haute précision (microkératome mécanique) et l'autre utilise des ondes infrarouges (laser femtoseconde).
Qu'est‐ce qui a été étudié dans cette revue ?
La myopie est un trouble médical dans lequel les personnes peuvent voir clairement les objets proches mais les objets plus éloignés sont flous. En 2010, la myopie touchait environ deux milliards de personnes dans le monde. La myopie peut être traitée en utilisant soit des lunettes, des lentilles de contact, ou une intervention chirurgicale. La chirurgie de la myopie modifie la forme de la structure transparente de la partie antérieure de l'œil (cornée). Le LASIK est la procédure chirurgicale la plus courante utilisée pour corriger la myopie.
La procédure LASIK crée un volet dans la cornée afin de la remodeler. Ce volet peut être réalisé soit par un microkératome mécanique, soit par un laser femtoseconde. Le microkératome mécanique utilise une lame pour faire le volet et le laser femtoseconde utilise un laser pour faire le volet. Les volets fabriqués par les deux méthodes sont différents en termes d'épaisseur et de structure, et les effets secondaires résultant de chaque méthode sont également différents. Nous avons recueilli et analysé toutes les études pertinentes pour répondre à cette question. Nous avons trouvé 11 études qui ont inclus 943 participants (1691 yeux).
Quels ont été les principaux résultats ?
Il n'y a pas de données probantes indiquant une différence dans les résultats de la vision entre l'utilisation d'un microkératome mécanique ou d'un laser femtoseconde. Les données probantes sont d’un niveau de confiance faible. Toutefois, il peut y avoir une différence d'effets secondaires entre les deux méthodes. Le groupe des microkératomes mécaniques a eu plus de cas de sécheresse oculaire et le groupe des lasers femtoseconde a eu plus de cas de gonflement de la cornée. Dans l'ensemble, en raison des données probantes d’un niveau de confiance faible, il est difficile de tirer une conclusion générale concernant l'efficacité et la sécurité de ces deux outils.
Cette revue est‐elle à jour ?
Les auteurs ont recherché les études qui avaient été publiés jusqu'au 22 février 2019.
Authors' conclusions
Summary of findings
| Laser‐assisted in‐situ keratomileusis (LASIK) with a mechanical microkeratome compared to LASIK with a femtosecond laser for LASIK in adults with myopia or myopic astigmatism | |||||
| Patient or population: adults (18 years) with more than 0.5 diopters of myopia or myopic astigmatism | |||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | |
|---|---|---|---|---|---|
| Risk with LASIK with a femtosecond laser | Risk with LASIK with a mechanical microkeratome | ||||
| Mean UCVA, 12 months after surgery (unit LogMAR) | The mean uncorrected visual acuity after 12 months surgery in the LASIK with a femtosecond laser group was ‐0.04 | MD 0.01 lower | ‐ | 42 | ⊕⊕⊝⊝ |
| Proportion of eyes within ± 0.5 diopters of target refraction, 12 months after surgery (unit diopters) | 925 per 1000 | 897 per 1000 | RR 0.97 | 79 | ⊕⊕⊝⊝ |
| Mean spherical equivalent of the refractive error, 12 months after surgery (unit diopters) | The mean spherical equivalent of the refractive error 12 months after surgery in the LASIK with a femtosecond laser group ranged from ‐0.30 to ‐0.31 | MD 0.09 higher | ‐ | 168 | ⊕⊕⊝⊝ |
| Corneal haze, any time point | 23 per 1000 | 8 per 1000 | RR 0.33 | 86 | ⊕⊝⊝⊝ |
| Dry eye, any time point | 80 per 1000 | 457 per 1000 | RR 5.74 | 183 | ⊕⊕⊝⊝ |
| Diffuse lamellar keratitis, any time point | 209 per 1000 | 56 per 1000 | RR 0.27 | 134 | ⊕⊕⊝⊝ |
| Epithelial ingrowth, any time point | 20 per 1000 | 20 per 1000 | RR 1.04 | 102 | ⊕⊝⊝⊝ |
| *The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| aDowngraded one level due to imprecision. | |||||
Background
Description of the condition
Refractive errors are an important cause of vision impairment and blindness (Holden 2016). Myopia (nearsightedness) is a type of refractive error that causes blurry vision at distance because abnormal structural conditions of the cornea, lens, or length of the eye prevent the images from focusing properly on the retina (Riordan‐Eva 2011). Myopia affected 1.95 billion people worldwide in 2010 (Holden 2016), with a higher prevalence in urban areas (Morgan 2012). The global economic burden generated by myopia has been estimated at USD 202 billion per annum (Smith 2009), and approximately USD 139 billion in the USA alone (NASEM 2016).
Description of the intervention
Refractive errors can be corrected through non‐surgical (eyeglasses and contact lenses) and surgical methods. The surgical methods are long‐lasting treatments that are used when a person becomes intolerant to contact lenses, encounters visual aberration from high‐powered spectacles, or desires to eliminate or reduce their dependence on glasses or contact lenses (Azar 2002). The eye is an optical system with a refractive power that can be changed by altering the curvature of its refractive surface or the location of elements of the system. Intraocular implants, such as intraocular lenses, can be used to correct refractive errors; however, the most common refractive surgeries performed in the USA are keratorefractive techniques. Keratorefractive surgeries are a group of techniques that modify the refractive power of the cornea by changing its curvature. These techniques include photorefractive keratectomy, laser subepithelial keratomileusis, intrastromal lenticule extraction, and laser‐assisted in‐situ keratomileusis (LASIK) (Bower 2001).
Due its safety and efficacy profile, LASIK is more popular compared with other surgeries. This technique creates a flap of the outermost parts of the cornea (epithelium, bowman layer, and anterior stroma) to expose the middle part of the cornea (stromal bed) and reshape it with excimer laser using photoablation (Ang 2009). LASIK comprises the creation of a corneal flap with an intended diameter that ranges from 7.8 mm to 9.8 mm and a thickness of 90 μm to 180 µm. The flap can be achieved by a mechanical microkeratome or a femtosecond laser (Farjo 2013). The mechanical microkeratome uses an oscillating blade to create the corneal flap (Bower 2001), while the femtosecond laser creates the flap with a focusable photodisruptive laser that delivers ultrashort (10–15 seconds) pulses with a wavelength within the infrared spectrum in a preset pattern (Lubatschowski 2000). This laser ionizes the tissue, causing molecular disruption within the cornea. The beam is focused on a small spot, creating electrically charged particles through multiphotonic absorption which release electrons from the atoms by a process known as avalanche ionization (Azar 2006). The free electrons transfer their energy to the surrounding medium, evaporating the adjacent tissue and forming cavitation bubbles consisting of carbon dioxide, nitrogen, and water (Bashir 2017). When the cavitation bubbles expand, they produce a regular and precise dissection of the corneal flap (Farjo 2013; Huhtala 2016; Sales 2016). The main differences between femtosecond and microkeratome flaps are the thickness and architecture. This Cochrane Review focused on the use of femtosecond laser or mechanical microkeratome in LASIK to correct myopia.
How the intervention might work
The femtosecond laser has some theoretical advantages over the use of a mechanical microkeratome in LASIK. For instance, the thickness of flaps created with the mechanical microkeratome have a significant variation (25 μm to 250 µm) compared with the femtosecond laser (78 μm to 173 µm). The predictability of the procedure to create a flap could be an important factor in the biomechanical integrity of the cornea (Flanagan 2003). Complications associated with the mechanical microkeratome are free or incomplete flaps, buttonholes, and epithelial erosions (Azar 2006). These complications are less common with the femtosecond laser, because it dissects in patterns that allow variation of flap width, depth, and diameter that may lead to better surgical results (Ang 2009; Gil‐Cazorla 2011; Issa 2011; Medeiros 2011). Other proposed benefits of the femtosecond laser use in LASIK are better uncorrected visual acuity (UCVA) (Gil‐Cazorla 2011), lower intraocular pressure during the procedure (Chaurasia 2010), and a lower incidence of dry eye (Salomão 2009).
The creation of the corneal flap with a mechanical microkeratome is free of some complications specific to LASIK with femtosecond laser, such as vertical gas breakthrough, anterior chamber bubbles, and opaque bubble layer (Azar 2006; Courtin 2015; Gatinel 2013; Moshirfar 2010; Stonecipher 2006). Another advantage of the mechanical microkeratome over the femtosecond laser is that it is fully reusable after dismantling and sterilization (the blade is the only single use part of the device), and there is no need to switch the patient to the excimer laser; these factors lower the overall cost of the procedure (Azar 2019). Other reported benefits of the mechanical microkeratome are lower risks of developing corneal haze (Patel 2008), transient light sensitivity (Stonecipher 2006), diffuse lamellar keratitis (Moshirfar 2010), and rainbow glare (Gatinel 2013).
Why it is important to do this review
The number of people with myopia globally is projected to increase to 4.76 billion by 2050 (Holden 2016). The trend has important economic (Corcoran 2015) and public health implications (NASEM 2016). LASIK is one of the most common surgical procedures used to correct refractive errors (Bower 2001). Femtosecond laser technology are often used in high‐income countries, while the mechanical microkeratome are used in low‐income countries (Salomao 2010). Both procedures have some advantages and disadvantages over the other, some of which are controversial (Ang 2009; Azar 2006; Bashir 2017; Farjo 2013). Therefore, it is important to evaluate the comparative effectiveness and safety of both procedure for myopia.
Objectives
To compare the effectiveness and safety of mechanical microkeratome versus femtosecond laser in LASIK for myopia.
Methods
Criteria for considering studies for this review
Types of studies
We included randomized controlled trials (RCTs). We did not restrict trial inclusion based on language or publication status.
Types of participants
People aged 18 years or older with more than 0.5 diopters of myopia or myopic astigmatism.
Types of interventions
We included trials that compared mechanical microkeratome with femtosecond laser in LASIK.
Types of outcome measures
Primary outcomes
-
Mean uncorrected visual acuity (UCVA) at 12 months after surgery. We used logMAR for visual acuity analyses.
Secondary outcomes
-
Mean UCVA at one and three months after surgery
-
Mean best corrected visual acuity (BCVA) at one, three, and 12 months after surgery
-
Proportion of eyes within ± 0.5 diopters of target refraction at one and 12 months after surgery
-
Proportion of eyes with loss of two or more lines of BCVA at 12 months after surgery
-
Mean spherical equivalent of the refractive error, measured in diopters, at one and 12 months after surgery
-
Intraoperative and postoperative pain at one day and one week, assessed with any validated measurement scale
-
Quality of life measures, assessed with any validated measurement scale at any point within follow‐up
Adverse outcomes
We considered adverse outcomes as reported by included trials up to 12 months after surgery. Specific adverse outcomes of interest were the following.
-
Corneal haze
-
Dry eye
-
Visual symptoms (double images, glares, halos, starburst)
-
Flap displacement
-
Flap melt
-
Diffuse lamellar keratitis
-
Infectious keratitis
-
Epithelial ingrowth
-
Corneal ectasia
Search methods for identification of studies
Electronic searches
The Cochrane Eyes and Vision Information Specialist (Lori Rosman) searched the following electronic databases for randomized controlled trials. There were no restrictions on language or year of publication. The electronic databases were last searched on 22 February 2019.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2019, Issue 2) (which contains the Cochrane Eyes and Vision Trials Register) in the Cochrane Library (searched 22 February 2019; Appendix 1).
-
MEDLINE Ovid (1946 to 22 February 2019; Appendix 2).
-
Embase.com (1947 to 22 February 2019; Appendix 3).
-
PubMed (1948 to 22 February 2019; Appendix 4).
-
LILACS (Latin American and Caribbean Health Science Information Database (1982 to 22 February 2019; Appendix 5).
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicalTrials.gov; searched 22 February 2019; Appendix 6).
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp; searched 22 February 2019; Appendix 7).
Searching other resources
We searched the reference lists of reports from trials we included in the review to look for additional trials. We did not conduct manual searches of conference proceedings or abstracts specifically for this review.
Data collection and analysis
Selection of studies
Two review authors (NKL, AN) assessed the titles and abstracts of articles identified through the literature search against inclusion criteria (listed in the Criteria for considering studies for this review section) and independently classified these as 'definitely relevant', 'possibly relevant', or 'definitely not relevant'. We used Covidence software to manage the screening process (Covidence). Any disagreement was resolved by a third review author (EGH). We obtained the full‐text copies of all trials classified as 'definitely relevant' or 'possibly relevant'. Each review author independently assessed each trial for inclusion and labeled it as either 'include' or 'exclude'. We contacted the authors of the primary trials via email for clarification whenever necessary. If there was no response within three weeks, we assessed the trial based on the information available. A third review author (EGH) resolved any disagreement. We documented the reason for exclusion of each trial excluded after reviewing the full report in a Characteristics of excluded studies table. We used Google Translate to assess trials written in languages other than English and Spanish. We used a PRISMA flowchart to depicts the flow of information through the different phases of the systematic review (Moher 2009).
Data extraction and management
Two review authors (NKL, AN) independently extracted data from the included trials using data extraction forms developed by Cochrane Eyes and Vision and accessed via Covidence (Appendix 8). A third review author (EGH) resolved any disagreements. We contacted authors of the primary trials via email to obtain missing information or to clarify data. We waited three weeks for a response; in the absence of a response, we used the available information, as provided in published reports. One review author (NKL) entered data into Review Manager 5 (Review Manager 2014), and a second review author (AN) verified the data entered.
Assessment of risk of bias in included studies
Two review authors (NKL, AN) evaluated the risk of selection (random sequence generation and allocation concealment before randomization), performance (masking of trial participants and personnel), detection (masking of outcome assessors), attrition (missing data and absence of an intention‐to‐treat analysis), reporting (selective outcome reporting), and other potential sources of bias using the Cochrane 'Risk of bias' assessment tool as described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We classified the risk of bias as 'low', 'high', or 'unclear' (insufficient information for assessment). We contacted authors of the primary trials when methods were unclear or when additional information about trial design or methods was required to assess the risk of bias. We waited three weeks for a response; in the absence of a response, we assessed the risk of bias based on descriptions provided in published reports. A third review author (EGH) resolved any disagreement between review authors.
Measures of treatment effect
We calculated mean differences (MDs) with 95% confidence intervals (CIs) for continuous outcomes such as mean UCVA after surgery, BCVA after surgery, and mean spherical equivalent of refractive error after surgery. For dichotomous outcomes (e.g. proportion of eyes within 0.5 diopters of target refraction after surgery), we calculated risk ratios (RRs) with 95% CIs. We used logMAR (logarithm of the minimum angle of resolution) for the outcomes that included visual acuity.
Unit of analysis issues
The participant was the primary unit of analysis whenever: only one eye per participant was enrolled in the trial; or two eyes of a participant were treated as a single unit after being administered the same treatment (e.g. mean values, binocular visual acuity). If neither of these conditions was met, the eye was considered the unit of analysis. If two eyes of the same participant were randomized to two different interventions, the correlation between the two eyes must be accounted for in the analysis; and in such case, we would only use the results that have accounted for this correlation.
Dealing with missing data
We contacted trial authors to obtain missing data or data reported unclearly in the trial reports. We allowed three weeks for trial authors to respond and used the available information whenever there was no response. We did not impute missing participant data for analysis.
Assessment of heterogeneity
We compared the participant characteristics, trial interventions, and outcomes across trials to assess for clinical and methodological heterogeneity. We used a visual inspection of forest plots and Chi² test statistics to assess statistical heterogeneity among estimates of effect size from the included trials. We used the I² statistic, which estimates the proportion of variation in observed effects not due to chance, to identify inconsistency among trials; an I² statistic value greater than 50% represented substantial heterogeneity (Higgins 2017).
Assessment of reporting biases
We did not perform a meta‐analysis with 10 or more trials, therefore a visual inspection of funnel plots of the intervention effect estimates for evidence of asymmetry was not meaningful. An asymmetric funnel plot may suggest small‐study effects, which could be the result of reporting bias, heterogeneity, or differences in the methodological quality of trials. We assessed selective outcome reporting as part of the 'Risk of bias' assessment among individual trials.
Data synthesis
We combined the effect estimates from individual trials using the random‐effects model when trials had no substantial clinical or methodological heterogeneity.
Subgroup analysis and investigation of heterogeneity
There were insufficient data from the trials to examine findings by the degree of myopia at baseline among the trial participants: low to moderate myopia (less than 6.0 diopters) and high myopia (6.0 diopters or more).
Sensitivity analysis
We performed sensitivity analyses for primary and secondary outcomes to explore the effects of restricting our analyses to trials in which the authors received no financial compensation from the industry. We also performed a sensitivity analysis for the trials that had at least 80% follow‐up of participants in each group in the trials with high risk of bias and to explore the effects of fixed‐effect versus random‐effects meta‐analyses. Post hoc, we planned to conduct a sensitivity analysis to examine the effects of restricting our analyses to trials for which the unit of analysis was the patient (one eye per participant) or trials that used paired‐eye design and accounted for correlation between the two eyes in their analysis. However, we did not perform this because the only meta‐analysis performed involved only two trials, both of which were paired‐eye design and did not report whether they accounted for correlation between the two eyes in their analysis.
Summary of findings
We summarized the findings of the review using the GRADE approach to assess the strengths and limitations of evidence for both primary and secondary outcomes using the GRADEpro Guideline Development Tool (GDT) software (GRADEpro 2015). Using this approach, we classified the evidence for each outcome as 'high', 'moderate', 'low', or 'very low' and documented reasons for our judgments. We included the following seven outcomes in summary of findings Table 1.
-
Mean UCVA, 12 months after surgery
-
Proportion of eyes within ± 0.5 diopters of target refraction, 12 months after surgery
-
Mean spherical equivalent of the refractive error, 12 months after surgery
-
Corneal haze, at any time point
-
Dry eye, at any time point
-
Diffuse lamellar keratitis, at any time point
-
Epithelial ingrowth, at any time point
In our protocol, we planned to include the following outcomes in this review, but there were insufficient data to include them in the 'Summary of findings' table.
-
Proportion of eyes with loss of two or more lines of BCVA, 12 months after surgery
-
Postoperative pain, within one week after surgery
-
Quality of life score, 12 months after surgery
Results
Description of studies
Results of the search
As of February 2019, the electronic searches resulted in 6560 titles and abstracts (Figure 1). After removal of duplicates, we screened 4802 records from which we identified 74 trials for full‐text review. We then excluded 57 records after full‐text review because they were either non‐randomized trials (56 records) or did not compare a femtosecond laser with a mechanical microkeratome (1 record). We classified one study as ongoing. There were five records from the Patel_group 2010 and two records from the Manche_group 2008. Overall, we included 16 records from 11 trials that met the inclusion criteria for the review (see Characteristics of included studies table).
Trial flow diagram.
Included studies
Participant selection
All included trials enrolled adults (aged > 18 years old) with spherical or spherocylindrical myopia who were suitable candidates for Laser‐assisted in‐situ keratomileusis (LASIK) after examination (Buzzonetti 2008; Durrie 2005; Gui‐Hong 2018; Hasimoto 2013; Manche_group 2008; Patel_group 2010; Salomão 2009; Tan 2007; Tran 2005; Zhai 2013; Zhou 2012).There were 943 participants (1691 eyes) enrolled across all trials; 547 participants (824 eyes) received LASIK with a mechanical microkeratome and 588 participants (867 eyes) received LASIK with a femtosecond laser. Each trial included between 9 and 360 participants. Six trials randomized one eye to one intervention and the fellow eye to another (Durrie 2005; Hasimoto 2013; Manche_group 2008; Patel_group 2010; Tan 2007; Tran 2005). The trials were conducted in the USA (5 trials), China (3 trials), and one each conducted in Brazil, Italy and Singapore. Preoperative refraction of participants was up to 7.50 diopters of sphere and 3 diopters of cylinder. The year of publication ranged from 2005 to 2018, with follow‐up time between one day and five years.
Interventions
Trials included in this review assigned participants to receive LASIK with a mechanical microkeratome or a femtosecond laser. One trial randomized 81 participants (161 eyes) into three intervention groups (Zhai 2013), of which two of the three intervention arms involving 60 participants (117 eyes), were relevant to this review. Three trials did not report review specific outcomes of interest (Tan 2007; Zhai 2013; Zhou 2012), therefore we did not include them in the analysis. The remaining eight trials included 482 participants (772 eyes) enrolled across all trials; 270 participants (365 eyes) received LASIK with a mechanical microkeratome and 312 participants (407 eyes) with a femtosecond laser. Among the eight trials that reported review specific outcomes of interest, two trials used a parallel‐trial design (Buzzonetti 2008; Gui‐Hong 2018), and six trials used a within‐person trial design (Durrie 2005; Hasimoto 2013; Manche_group 2008; Patel_group 2010; Salomão 2009; Tran 2005). All eight trials were similar in treatment setting.
Outcomes
Three trials reported the mean uncorrected visual acuity (UCVA) after surgery (Durrie 2005; Gui‐Hong 2018; Patel_group 2010). One trial reported best corrected visual acuity (BCVA) after surgery (Patel_group 2010). Five trials reported the mean spherical equivalent of the refractive error after surgery (Buzzonetti 2008; Durrie 2005; Gui‐Hong 2018; Manche_group 2008; Patel_group 2010), two of these trials reported proportion of eyes within 0.5 diopters of target refraction after surgery (Durrie 2005; Manche_group 2008). Four trials reported adverse events (Hasimoto 2013; Manche_group 2008; Salomão 2009; Tran 2005). We considered the methods used to measure the outcomes to be consistent across trials. None of the trials assessed proportion of eyes with loss of two or more lines of BCVA from preoperative visual acuity, pain, quality of life, visual symptoms, flap displacement, flap melt, infectious keratitis, or corneal ectasia.
Funding sources
Four trials reported source of funding. One trial was funded by the IRCCS‐Casa Sollievo della Sofferenza Hospital and Institue of Ophthalmology of Catholic University (Buzzonetti 2008), another was funded by the National Institutes of Health, Research to Prevent Blindness, and the Mayo Foundation (Patel_group 2010). Salomão 2009 was supported by US Public Health Service grants from the National Eye Institute, and Research to Prevent Blindness. IntraLase Corp funded one trial (Tran 2005), and the remaining seven trials did not report funding sources.
Ongoing Study
We identified one ongoing trial (PACTR201708002498199).
Excluded studies
We reviewed and excluded 57 records. We described the reasons for exclusion in the Characteristics of excluded studies table.
Risk of bias in included studies
A summary of the risk of bias for included trials is presented in Figure 2 and a risk of bias graph is presented in Figure 3.
Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included trials
Allocation
We judged four trials at low risk of bias for random sequence generation because the trial investigators described a random component in the sequence generation process (Durrie 2005; Hasimoto 2013; Manche_group 2008; Zhou 2012). The remaining seven trials did not specify the method for randomization, therefore, we judged the risk to be unclear (Buzzonetti 2008; Gui‐Hong 2018; Patel_group 2010; Salomão 2009; Tan 2007; Tran 2005; Zhai 2013).
Two trials reported using sealed and opaque envelopes to conceal treatment allocation (Hasimoto 2013; Manche_group 2008), and we judged them at low risk of bias. The remaining nine trials did not describe the method of allocation concealment and we judged them at unclear risk of bias
Blinding
One trial specified masking of both participants and trial personnel (Manche_group 2008), and another reported masking of participants (Hasimoto 2013), and we judged them at low risk of bias. Patel_group 2010 stated that "it was not possible to mask patients…", therefore we judged this trial at high risk of bias. The remaining trials included in the review did not specify how masking of participants and personnel was achieved and we judged them at unclear risk of bias.
Three trials described adequate masking of the outcome assessors (Durrie 2005; Hasimoto 2013; Patel_group 2010), and we judged them at low risk of detection bias. The remaining eight trials did not provide a clear description of the measures used to mask outcome assessors or did not address this issue, therefore, we judged them at unclear risk of bias.
Incomplete outcome data
We judged five trials at low risk of bias; Patel_group 2010, Durrie 2005, and Salomão 2009 reported no missing outcome data. Manche_group 2008 described missing outcome data that we considered balanced in numbers across intervention groups with similar reasons for missing data across the two groups. Tran 2005 reported one amblyopic eye that was dropped from the analysis; we considered that reason for missing data unlikely to be related to the outcomes being explored in the review.
Hasimoto 2013 reported missing outcome data of three participants (6 eyes), one participant (2 eyes) had a head trauma (intervention group was not specified), another participant (1 eye) presented with diffuse lamellar keratitis (femtosecond laser group), and one other participant (1 eye) presented with bleeding from limbal vessels (intervention group was not specified). We considered the missing outcome data likely to be related to true outcomes, therefore we judged this trial at high risk of bias.
The remaining five trials had insufficient reporting of incomplete outcome data and we judged them at unclear risk of bias.
Selective reporting
Outcomes reported in the different reports of the Patel_group 2010 were not prespecified in trial registration, therefore we judged it at unclear risk of bias. We also judged the remaining 10 trials at unclear risk of bias for selective reporting because we did not identify trial protocols.
Other potential sources of bias
Two trials declared financial compensations from the industry directly related to one of the intervention groups (Durrie 2005; Tran 2005), and we judged them to be at high risk of other bias. The remaining trials appear to be free from other sources of bias, and we judged them at low risk for other bias.
Effects of interventions
See: summary of findings Table 1
Mean uncorrected visual acuity (UCVA) after surgery
One trial provided data of 42 eyes for the primary outcome (Patel_group 2010). The mean difference (MD) of UCVA at 12 months after surgery was –0.01, (95% confidence interval (CI) –0.06 to 0.04). Three trials provided data from 192 eyes for the mean UCVA at one month (Durrie 2005; Gui‐Hong 2018; Patel_group 2010) (MD 0.13, ‐0.08 to 0.33). Two trials provided data of mean UCVA at three months after surgery (Durrie 2005; Patel_group 2010). There was low statistical heterogeneity between the two trials, the I² statistic was 0% and the Chi² test for heterogeneity was not statistically significant (P = 1.00). Therefore, we combined the data of UCVA at three months in a meta‐analysis (72 eyes). The observed MD of UCVA was 0.00 (95% CI –0.03 to 0.03; Analysis 1.1; Figure 4). We did not perform an overall UCVA analysis because two groups included data from the same trial and were based on two clinically different time points. The certainty of evidence for this outcome across the various trials was low; we downgraded for risk of bias and imprecision.
![Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.1 Mean uncorrected visual acuity after surgery [logMAR].](/cdsr/doi/10.1002/14651858.CD012946.pub2/media/CDSR/CD012946/image_n/nCD012946-FIG-04.svg)
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.1 Mean uncorrected visual acuity after surgery [logMAR].
Proportion of eyes within ± 0.5 diopters of target refraction after surgery
Data from the trial by Manche_group 2008 comprised 79 eyes and found that the proportion of eyes 0.5 diopters within target refraction was 3% lower after LASIK with mechanical microkeratome 12 months after surgery: risk ratio (RR) 0.97, 95% CI 0.85 to 1.11; low‐certainty evidence; Analysis 1.2; we downgraded for risk of bias and imprecision. Two other trials with 125 participants (Durrie 2005; Patel_group 2010), reported the proportion of eyes within 0.5 diopters of target refraction after LASIK one month after surgery. We combined the data of the two trials because heterogeneity was low, the I² statistic was 0% and the Chi² test for heterogeneity was not statistically significant (P = 0.67). There was no difference between LASIK with a mechanical microkeratome or a femtosecond laser of achieving 0.5 diopters within target refraction one month after surgery (Analysis 1.2; Figure 5). We did not perform an overall analysis because the groups were based on two clinically different time points. However, the certainty of evidence for this outcome across the various time points and trials was low; we downgraded for risk of bias and imprecision.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.2 Proportion of eyes within ± 0.5 diopters of target refraction after surgery.
Mean spherical equivalent of the refractive error after surgery
Durrie 2005, Patel_group 2010, and Gui‐Hong 2018 reported mean spherical equivalent of refractive error one month after surgery. There was considerable heterogeneity (I² = 96%, Chi² test for heterogeneity was statistically significant, P < 0.0001) thus, rendered a summary of the pooled effect estimate inappropriate. Point estimates from the three trials showed no evidence of a difference between the two groups: Durrie 2005 (102 eyes, 51 participants); MD 0.40, 95% CI 0.29 to 0.51; Gui‐Hong 2018: (240 eyes, 120 participants); MD –0.01, 95% CI –0.05 to 0.03; Patel_group 2010: (42 eyes, 21 participants); MD –0.06, 95% CI –0.24 to 0.12; Analysis 1.3; Figure 6). However, data from Durrie 2005, which seem to be an outlier, indicate that participants undergoing femtosecond LASIK seem to do better in spherical equivalent of the refractive error. The possible explanation probably has to do with the flap thickness predictability and biomechanics of the corneal flap, which is more efficiently created with the femtosecond laser, and has low spherical aberrations compared to microkeratome LASIK (Bashir 2017). The certainty of evidence across the included trials was low; we downgraded for risk of bias and imprecision.
![Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.3 Mean spherical equivalent of the refractive error after surgery [diopters].](/cdsr/doi/10.1002/14651858.CD012946.pub2/media/CDSR/CD012946/image_n/nCD012946-FIG-06.svg)
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.3 Mean spherical equivalent of the refractive error after surgery [diopters].
Three trials reported mean spherical equivalent of refractive error 12 months after surgery (Buzzonetti 2008; Manche_group 2008; Patel_group 2010; 168 eyes, 92 participants). We combined the data of the three trials in a meta‐analysis and observed no evidence of a difference in mean spherical equivalent of refractive error between the two groups (MD 0.09, 95% CI –0.01 to 0.19; I² = 0%; Analysis 1.3; Figure 6). The certainty of evidence was low; we downgraded for risk of bias and imprecision.
Corneal haze
Based on the trial by Manche_group 2008 (86 eyes), the RR of corneal haze was 77% lower with the mechanical microkeratome group compared with the femtosecond laser group (RR 0.33, 95% CI 0.01 to 7.96; Figure 7). The certainty of evidence was very low; we downgraded for risk of bias and very serious imprecision.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.4 Corneal haze.
Dry eye
The trial by Salomão 2009 (183 eyes, 183 participants) had a RR of 5.74 for the participants assigned to the mechanical microkeratome group compared with the femtosecond laser group (95% CI 2.92 to 11.29; Figure 8). The certainty of evidence was low; we downgraded for risk of bias and imprecision. The trial by Manche_group 2008 described a self‐reported mean dry eye score at preoperative, and one, three, six, and 12 months postoperative time points. We did not include the trial in the quantitative synthesis because 100% of the participants reported dry eye preoperatively. The findings reflected an increase in the mean dry eye score postoperatively in the femtosecond laser group one month after surgery (P = 0.02), the score returned to baseline by six months after surgery. The certainty of evidence was low; we downgraded for risk of bias and imprecision.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.5 Dry eye.
Diffuse lamellar keratitis
Manche_group 2008, Hasimoto 2013, and Tran 2005 (134 eyes, 67 participants) reported diffuse lamellar keratitis. We combined the data of the trials in a meta‐analysis since there was no heterogeneity (I² = 0%, Chi² P = 0.89). The RR for the participants assigned to the mechanical microkeratome group was 73% lower compared with the femtosecond laser group (RR 0.27, 95% CI 0.10 to 0.78; Analysis 1.6; Figure 9). The certainty of evidence was low; we downgraded for risk of bias and imprecision.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.6 Diffuse lamellar keratitis.
Epithelial ingrowth
Two trials reported epithelial ingrowth (Manche_group 2008; Tran 2005; 102 eyes, 51 participants). We combined data in a meta‐analysis and observed a RR of 1.04 for the participants assigned to the mechanical microkeratome group compared with the femtosecond laser group (95% CI 0.11 to 9.42; I² = 0%; Analysis 1.7; Figure 10). The certainty of evidence was very low; we downgraded for risk of bias and very serious imprecision.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.7 Epithelial ingrowth.
Sensitivity analysis
The results of the sensitivity analyses of the trials in which the authors received financial compensation from the industry, trials that had at least 80% follow‐up of participants in each group, or trials with high risk of bias, or the use of fixed‐effect versus random‐effects meta‐analysis showed no difference in the findings (data not shown).
Discussion
Summary of main results
We included 16 reports of 11 trials in this review representing 943 participants (1691 eyes). The trials allocated adults 18 years and older, with spherical or spherocylindrical myopia considered suitable candidates for Laser‐assisted in‐situ keratomileusis (LASIK) after examination, to a mechanical microkeratome group or a femtosecond laser group. Based on our assessment of the trials that reported uncorrected visual acuity (UCVA), there was no evidence of a difference in UCVA after LASIK in eyes with the flap created by a femtosecond laser compared to eyes with the flap created by a mechanical microkeratome. There was also no evidence of a difference between the two treatments in terms of achieving within 0.5 diopters of target refraction or in mean spherical equivalent of refractive error 12 months after surgery. The association between corneal haze and epithelial ingrowth with the interventions was uncertain. We observed a lower risk of diffuse lamellar keratitis in the mechanical microkeratome group compared with the femtosecond laser group when combining the results of three trials. Only one trial reported dry eye after the intervention. This trial showed that there may be a higher risk of dry eye in the mechanical microkeratome compared to the femtosecond laser group. We found no trials that reported proportion of eyes with loss of two or more lines of best corrected visual acuity (BCVA) from preoperative visual acuity, pain, quality of life, visual symptoms, flap displacement, flap melt, infectious keratitis, or corneal ectasia. These findings were robust to our sensitivity analysis.
Overall completeness and applicability of evidence
The trials included in the review were not substantially heterogeneous from a clinical point of view. We considered that the trials identified were not sufficient to address all the objectives of the review. For instance, relevant outcomes such as, quality of life, proportion of eyes with loss of two or more lines of BCVA from preoperative visual acuity, as well as intraoperative and postoperative pain should have been investigated.
Potential biases in the review process
Two review authors independently assessed the titles and abstracts and full‐text reports of potentially eligible trials; a third review author resolved any conflicts. The extraction of data and assessment of risk of bias for all trials included in this review were done in the same manner. During the review process, we changed four outcomes in the 'Summary of findings' table specified in the protocol due to insufficient data available from the trials. All outcomes incorporated in summary of findings Table 1 were prespecified secondary outcomes in the protocol. We are aware of no other potential biases in the review process.
Certainty of the evidence
We graded five outcomes reported in summary of findings Table 1 at low‐certainty evidence and two at very low‐certainty evidence. All the trials that reported the seven outcomes had serious risk of bias and we downgraded them. We also downgraded the certainty of evidence of all the adverse events outcomes due to serious or very serious concerns related to imprecision in the results. We identified trial registration in only one trial and there were inconsistencies between trial outcomes defined in the protocol and those reported in the trial publications. Two trials were sponsored by companies related to one of the interventions under investigation and had roles in the design or analysis of these trials (Durrie 2005; Tran 2005). Some investigators reported financial relationships with the company.
Agreements and disagreements with other studies or reviews
Chen 2012 published one systematic review and meta‐analysis comparing IntraLase femtosecond laser versus mechanical microkeratomes in LASIK for myopia. The purpose of the review was to evaluate the safety, efficacy, and predictability of the two interventions. The trials included were randomized and non‐randomized trials. The primary outcomes were loss of more than two lines of corrected distance visual acuity, uncorrected distance visual acuity 20/20 or better, manifest refraction spherical equivalent within 0.50 diopters, final refractive spherical equivalent, and astigmatism. The secondary outcomes were flap thickness predictability, changes in higher‐order aberrations, and complications.
The trial investigators concluded the results showed no differences in visual acuity‐related outcomes between the IntraLase group and the mechanical microkeratomes group. These findings are consistent with the results of this review. Regarding complications, Chen 2012 reported higher events of diffuse lamellar keratitis in the femtosecond laser group, higher risk of loose epithelium in one or two quadrants in the group of mechanical microkeratome, and no significant difference in epithelial ingrowth between groups. Our results were consistent with a higher risk of diffuse lamellar keratitis in the femtosecond laser group as we observed a RR that was 73% lower in the mechanical microkeratome group compared with the femtosecond laser group (Analysis 1.6; Figure 9). Concerning intraoperative epithelial defects, we did not include this adverse event in our prespecified outcomes, therefore, we did not include it in the review. The findings of Chen 2012 regarding epithelial ingrowth were consistent with those presented in this review (Figure 10).
Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included trials
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.1 Mean uncorrected visual acuity after surgery [logMAR].
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.2 Proportion of eyes within ± 0.5 diopters of target refraction after surgery.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.3 Mean spherical equivalent of the refractive error after surgery [diopters].
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.4 Corneal haze.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.5 Dry eye.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.6 Diffuse lamellar keratitis.
Forest plot of comparison: 1 LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, outcome: 1.7 Epithelial ingrowth.
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 1: Mean uncorrected visual acuity after surgery
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 2: Proportion of eyes within ± 0.5 diopters of target refraction after surgery
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 3: Mean spherical equivalent of the refractive error after surgery
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 4: Corneal haze
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 5: Dry eye
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 6: Diffuse lamellar keratitis
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 7: Epithelial ingrowth
Comparison 1: LASIK with a mechanical microkeratome versus LASIK with a femtosecond laser, Outcome 8: Best corrected visual acuity after surgery
| Laser‐assisted in‐situ keratomileusis (LASIK) with a mechanical microkeratome compared to LASIK with a femtosecond laser for LASIK in adults with myopia or myopic astigmatism | |||||
| Patient or population: adults (18 years) with more than 0.5 diopters of myopia or myopic astigmatism | |||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | |
|---|---|---|---|---|---|
| Risk with LASIK with a femtosecond laser | Risk with LASIK with a mechanical microkeratome | ||||
| Mean UCVA, 12 months after surgery (unit LogMAR) | The mean uncorrected visual acuity after 12 months surgery in the LASIK with a femtosecond laser group was ‐0.04 | MD 0.01 lower | ‐ | 42 | ⊕⊕⊝⊝ |
| Proportion of eyes within ± 0.5 diopters of target refraction, 12 months after surgery (unit diopters) | 925 per 1000 | 897 per 1000 | RR 0.97 | 79 | ⊕⊕⊝⊝ |
| Mean spherical equivalent of the refractive error, 12 months after surgery (unit diopters) | The mean spherical equivalent of the refractive error 12 months after surgery in the LASIK with a femtosecond laser group ranged from ‐0.30 to ‐0.31 | MD 0.09 higher | ‐ | 168 | ⊕⊕⊝⊝ |
| Corneal haze, any time point | 23 per 1000 | 8 per 1000 | RR 0.33 | 86 | ⊕⊝⊝⊝ |
| Dry eye, any time point | 80 per 1000 | 457 per 1000 | RR 5.74 | 183 | ⊕⊕⊝⊝ |
| Diffuse lamellar keratitis, any time point | 209 per 1000 | 56 per 1000 | RR 0.27 | 134 | ⊕⊕⊝⊝ |
| Epithelial ingrowth, any time point | 20 per 1000 | 20 per 1000 | RR 1.04 | 102 | ⊕⊝⊝⊝ |
| *The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| aDowngraded one level due to imprecision. | |||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Mean uncorrected visual acuity after surgery Show forest plot | 3 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.1.1 At 1 month | 3 | 384 | Mean Difference (IV, Random, 95% CI) | 0.13 [‐0.08, 0.33] |
| 1.1.2 At 3 months | 2 | 144 | Mean Difference (IV, Random, 95% CI) | 0.00 [‐0.03, 0.03] |
| 1.1.3 At 12 months | 1 | 42 | Mean Difference (IV, Random, 95% CI) | ‐0.01 [‐0.06, 0.04] |
| 1.2 Proportion of eyes within ± 0.5 diopters of target refraction after surgery Show forest plot | 3 | Risk Ratio (IV, Random, 95% CI) | Subtotals only | |
| 1.2.1 At 1 month | 2 | 144 | Risk Ratio (IV, Random, 95% CI) | 0.87 [0.77, 0.99] |
| 1.2.2 At 12 months | 1 | 79 | Risk Ratio (IV, Random, 95% CI) | 0.97 [0.85, 1.11] |
| 1.3 Mean spherical equivalent of the refractive error after surgery Show forest plot | 5 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.3.1 At 1 month | 3 | 384 | Mean Difference (IV, Random, 95% CI) | 0.11 [‐0.18, 0.40] |
| 1.3.2 At 12 months | 3 | 168 | Mean Difference (IV, Random, 95% CI) | 0.09 [‐0.01, 0.19] |
| 1.4 Corneal haze Show forest plot | 1 | Risk Ratio (IV, Random, 95% CI) | Totals not selected | |
| 1.4.1 Any time point | 1 | Risk Ratio (IV, Random, 95% CI) | Totals not selected | |
| 1.5 Dry eye Show forest plot | 1 | Risk Ratio (IV, Random, 95% CI) | Totals not selected | |
| 1.5.1 Any time point | 1 | Risk Ratio (IV, Random, 95% CI) | Totals not selected | |
| 1.6 Diffuse lamellar keratitis Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.6.1 Any time point | 3 | 134 | Risk Ratio (M‐H, Random, 95% CI) | 0.27 [0.10, 0.78] |
| 1.7 Epithelial ingrowth Show forest plot | 2 | Risk Ratio (IV, Random, 95% CI) | Subtotals only | |
| 1.7.1 Any time point | 2 | 102 | Risk Ratio (IV, Random, 95% CI) | 1.04 [0.11, 9.42] |
| 1.8 Best corrected visual acuity after surgery Show forest plot | 1 | 126 | Mean Difference (IV, Fixed, 95% CI) | ‐0.01 [‐0.04, 0.02] |
| 1.8.1 At 1 month | 1 | 42 | Mean Difference (IV, Fixed, 95% CI) | 0.00 [‐0.05, 0.05] |
| 1.8.2 At 3 months | 1 | 42 | Mean Difference (IV, Fixed, 95% CI) | 0.00 [‐0.05, 0.05] |
| 1.8.3 At 12 months | 1 | 42 | Mean Difference (IV, Fixed, 95% CI) | ‐0.03 [‐0.09, 0.03] |
