Main

Phacoemulsification and foldable intraocular lens (IOL) implantation has led to improved success rates and faster visual rehabilitation in patients undergoing cataract surgery. The refractive outcome following phacoemulsification cataract surgery is dependent on a number of factors.1,2,3 They include axial length measurement, keratometry, anterior chamber depth, IOL power formulae, and the quality of the IOL. Of these factors, inaccurate axial length measurements were shown to be the major deterrent to the predictability of the refractive outcome.4,5 Since the predictability of refractive outcome is based on the accuracy of preoperative biometry, the methods used in biometry continue to evolve.6,7,8,9,10,11,12

Preoperative biometry performed using A-scan ultrasonography uses the echo delay time to measure intraocular distances. It has a longitudinal resolution of 200 μm and an accuracy of 100–120 μm in measuring axial lengths.6,7 Studies have shown that an error of 100 μm in axial length measurement could lead to 0.28 D of postoperative refractive error.5 Further, the ultrasound technique requires contact with the eye for measuring the axial length and the applanation method suffers from the disadvantage of corneal indentation during measurement. Recently, optical coherence tomography has found its clinical application in preoperative biometry.11,12 This technique aims to improve the precision in axial length measurements using the principle of partial coherence laser interferometry (PCLI). A dual beam of infrared light (780 nm) of short coherence length (160 μm) with different optical lengths is emitted by a laser diode source. The eye to be measured and the photodetector are situated at each leg of the interferometer. Both partial beams are reflected at the corneal surface and the retina (RPE). Interference occurs if the path difference between the beams is smaller than the coherence length. The interference signal received by the photodetector is measured dependent on the position of the interferometer mirror, which could be measured precisely. This measurement gives the optical length between the corneal surface and retina. The optical distance is used to derive geometric intraocular distances by incorporating the group refractive indices of the respective ocular media (cornea, lens, aqueous and vitreous humor). This technique of optical biometry is reported to have a high resolution (12 μm) and precision (0.3–10 μm) in measuring intraocular distances as compared to conventional ultrasound.13,14

In this study we compared the refractive outcome in cataract surgery following biometry with the applanation A-scan ultrasound and partial coherence laser interferometry. The aim of the study was to evaluate the predictability of refractive outcome using optical and ultrasound biometry. The use of partial coherence laser interferometry in pseudophakic axial length measurements and its role in improving the refractive outcome was further analysed in this study.

Subjects and methods

In this prospective randomised clinical trial, 100 patients undergoing phacoemulsification cataract surgery were randomised to undergo biometry with either A-scan ultrasound or partial coherence laser interferometry (optical biometry). Fifty eyes of 50 patients underwent biometry with ultrasonography and the other 50 patients underwent partial coherence laser interferometry. Complicated cataracts related to chronic uveitis, trauma, or silicone oil were not included in the trial. The applanation A-scan, Nidek Echoscan-2000 was used for ultrasound biometry and the IOL Master (Zeiss Humphrey Systems) was used for partial coherence laser interferometry. Javal Schiotz keratometer was used for corneal curvature measurements for patients in the ultrasound group. The patients consented and the preoperative biometry was performed by an experienced biometrist on all the patients. The reliability of intraocular distance measurements was checked based on the sound to noise ratio (>2) in partial coherence laser interferometry and the retinal spikes in ultrasonography. The SRK-T formula was used to calculate the IOL power in all the patients. The A-constant was kept constant for all the eyes in this study. The desired postoperative refraction, based on the pre-existing refractive error was decided prior to surgery. The patients underwent phacoemulsification procedure through a 4.1 mm superior corneal tunnel and a folding IOL (Acrysof MA60BM, Alcon Lab, Dallas, TX, USA) was implanted by the same surgeon (JB). The patients were followed up on the first postoperative day, 1 week later and at 2 months by experienced observers (MR amd IK). The postoperative refraction was carried out at 2 months with an autorefractor and confirmed by subjective refraction. All patients in the study underwent pseudophakic axial length measurements by IOL Master at 2 months, irrespective of the type of biometry performed preoperatively. The postoperative axial length measurements were carried out by the same biometrist to avoid observer differences.

The postoperative mean spherical equivalent (MSE) was calculated for each of the patients and it was compared with the desired refraction. The mean error (ME) and mean absolute error (MAE) were derived based on the difference between the predicted and attained postoperative refraction. The postoperative refractive outcome was compared between the groups that underwent biometry with ultrasound technique as opposed to partial coherence laser interferometry. The preoperative axial length measurements performed by the ultrasound and partial coherence laser interferometry were compared with pseudophakic axial length measurements (IOL Master) using paired t-tests, to evaluate statistical significance (P < 0.05).

Results

The mean age of patients in the ultrasound group was 71 ± 8 yrs (range 40–86 yrs) and 67 ± 6 yrs (range 38–80 yrs) in the PCLI group. The mean preoperative axial length (AL) in the ultrasound group was 23.43 ± 1.2 mm and 23.47 ± 1.1 mm in the PCLI group (range 20.1–27 mm and 20.0–27.6 mm respectively). Forty-six of fifty patients underwent successful optical biometry in the PCLI group as compared to all 50 patients in the ultrasound group. Four eyes in the PCLI group had to undergo ultrasound biometry for AL measurements. All the eyes in the study had the IOL implanted in the capsular bag. The postoperative mean absolute error (MAE) was 0.6 ± 0.4 dioptres in patients who underwent ultrasound biometry (upper and lower 95% CI of 0.75 and 0.50 D respectively). The MAE in the PCLI group was 0.52 ± 0.35 D (upper and lower 95% CI of 0.63 and 0.42 respectively.) There was no statistically significant difference between the groups in terms of postoperative refractive outcome (P = 0.24, NS). Eighty per cent of the eyes in the ultrasound group achieved postoperative refraction of ±1 D of the predicted value as compared to 87% of patients in the PCLI group (Figure 1). On further analysis of the postoperative refraction, the eyes that underwent partial coherence laser interferometry had increased tendency for a hyperopic shift (65%), when compared to eyes in the ultrasound group (50%).

Figure 1
figure 1

The postoperative refraction of eyes that underwent optical biometry compared with ultrasound biometry. PCLI, Partial coherence laser interferometry. US, Ultrasound.

The pseudophakic axial length was measured using IOL Master by choosing the appropriate setting for the IOL material used in our study. The preoperative and postoperative AL were compared in both the groups. The mean postoperative AL in the ultrasound group was 23.54 ± 1.2 mm, the mean error 0.07 ± 0.2 mm and MAE of 0.19 ± 0.13 mm (P = 0.02). The eyes in the ultrasound group measured approximately 90–100 μm longer with partial coherence laser interferometry (Figure 2). The mean postoperative AL in the PCLI group was 23.35 ± 1.1 mm, the mean error −0.06 ± 0.17 mm and MAE of 0.13 ± 0.13 mm (P = 0.003). There was a statistically significant difference between the preoperative and postoperative AL values in both the groups. There was a mean shortening of the eyes in the PCLI group following cataract surgery (Figure 3). Ten eyes (20%) in the ultrasound group had postoperative refraction more than ±1 D. These pseudophakic eyes underwent retrospective biometry and IOL power calculations using partial coherence laser interferometry. The retrospective calculations revealed a 16% improvement in the predictability of postoperative refraction in these eyes.

Figure 2
figure 2

The difference in preoperative and postoperative axial length measurements in the ultrasound group.

Figure 3
figure 3

The difference in preoperative and postoperative axial length measurements in eyes that underwent optical biometry.

Four eyes (8%) failed partial coherence laser interferometry due to fixation instability related to macular degeneration (4%) and dense cataracts (4%). Ultrasound biometry was performed successfully using ultrasound technique in all four eyes that failed optical biometry (Table 1).

Table 1 Comparison between partial coherence laser interferometry and ultrasound biometry

Discussion

Our study compared the refractive outcome between applanation ultrasound and partial coherence laser interferometry. Both the groups compared favourably with no significant difference in functional outcome.15 However the patients who had partial coherence laser interferometry did better in reaching ±1 D of the expected post op refraction (87%). Our study has shown that partial coherence laser interferometry improves the predictive value for postoperative refraction by 16%, when compared to ultrasound biometry using retrospective IOL power calculations in pseudophakic eyes. Other studies have shown an improvement in predictive value up to 27%.16,17 It is possible to achieve better predictive value with PCLI by altering the A-constant. However we did not alter the A-constant in this study. Further, the study did not include eyes with axial lengths less than 20 mm and more than 28 mm. This could have an impact on the results and the predictive value. These are the limitations of this study.

Partial coherence laser interferometry is a non-contact method and offers the ease of obtaining keratometry values, anterior chamber depth and AL measurements in a single sitting. This is a significant advantage when compared to conventional ultrasound biometry, which demands topical anaesthesia for corneal applanation and is time consuming. Further, the precision achieved with partial coherence laser interferometry was shown to be 10 times better than that of ultrasound in earlier studies.16,17 The data from our study show that there is a tendency for hyperopic shift in eyes that undergo partial coherence laser interferometry. This is probably because the axial lengths are measured approximately 100 μm longer than with applanation ultrasound. This systematic error calls for caution while predicting postoperative refraction and we suggest appropriate alteration to the A-constants.

Since partial coherence laser interferometry relies on adequate foveal fixation, eyes with corneal scarring, dense cataracts, posterior capsule plaque, macular degeneration, and eccentric fixation fail to obtain reliable results. We report 8% failure rate in our study, which is comparable to other studies. However in areas where hard cataract predominates, the failure rate can be more and vice versa. On the other hand, PCLI has an edge over ultrasound biometry in measuring the AL of eyes with silicone oil or posterior staphyloma.

The disadvantages of using A-scan ultrasonography in pseudophakic eyes were reported in earlier studies.18 The role of PCLI in pseudophakic AL measurement was evaluated in this study. It revealed a mean shortening of the eyes postoperatively in the PCLI group. Earlier studies have failed to prove true shortening of the eyes following cataract surgery.18,19,20,21,22 The mean shortening encountered with PCLI is most likely related to the group refractive index incorporated in the calculation, causing this systematic error.23 The A constant will need to be altered to suit PCLI in order to achieve better accuracy.24 The partial coherence laser interferometry is able to achieve reliable AL measurements in pseudophakic eyes as observed in our study. This application becomes clinically relevant in evaluating pseudophakic eyes that might need a secondary piggyback IOL.

Thus the dual beam partial coherence laser interferometry improves the predictive value of postoperative refraction in eyes undergoing phacoemulsification cataract surgery. It is less time consuming and has the advantages of improved precision and patient acceptability when compared to conventional applanation ultrasound biometry. Partial coherence laser interferometry is reliable in pseudophakic axial length measurements when compared to ultrasound biometry.

Acknowledgments

The authors do not have any financial interest in the instruments described in the study.The authors would like to thank Steve Powell, Sheila Craft and Marion Oakley for their invaluable help with the study.