Aberrometry-Based Refraction During Cataract Surgery
Aberrometry-Based Refraction During Cataract Surgery
The study population encompassed 74 eyes (34 right and 40 left eyes) from 74 patients (42 male and 32 female). The mean age of the patients was 69.0 (SD 11.3) years, and the baseline manifest refraction was +0.31 sphere (SD 2.9) and −1.1 cylinder (SD 1.05).
Table 1 provides an overview of descriptive statistics on preoperative intraoperative and postoperative WA refractive data.
Figure 1A plots the median absolute differences of SE between neighbouring measurement points. The detailed differences between the preoperative measurement on the seated patient versus the first intraoperative measurement are shown in figure 1B. Repeated SE measurements with limits of agreements in aphakia and pseudophakia are shown in figure 2A and B. Test-retest reliability amounted to 0.98 and 0.88, respectively, which being >0.75 can be classified as good. Figure 3A and B plot the mean absolute differences of J0 and J45 between neighbouring measurement points (black circles).
(Enlarge Image)
Figure 1.
Absolute differences of spherical equivalent (SE). (A) Median absolute differences of SE between neighbouring measurement points (line within the box) with their respective quartiles (top and bottom of the box) and mean absolute differences (rhombus within the box, all ps<0.05). (B) Bland–Altman plot comparing preoperative SE measurements on the seated patient versus the first intraoperative measurement on the horizontal patient. Mean difference is shown by solid black line and limits of agreement by dashed lines. The 95% CIs for the lower and upper limits of agreement are −4.35 to −2.56 D and 3.62 to 1.83 D, respectively. Based on the sample size (N=38), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤ 0.9 D (±1.58). Measurement points M0-12 and differing Ns across those are explained in table 1.
(Enlarge Image)
Figure 2.
Spherical equivalent (SE) reproducibility in aphakia and pseudophakia; (A) Bland–Altman plot comparing threefold repeated SE measurements in aphakia. Mean difference is shown by solid black line and limits of agreement by dashed lines. The 95% CIs for the lower and upper limits of agreement are −0.96 to −0.47 D and 0.93 to 0.44 D, respectively. Based on the sample size (n=27), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤0.2 D (±0.36). (B) Bland–Altman plot comparing threefold repeated SE measurements in pseudophakia after removal of ophthalmic viscosurgical device (OVD). The 95% CIs for the lower and upper limits of agreement are −2.14 to −0.94 D and 2.13 to 0.93 D, respectively. Based on the sample size (n=22), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤0.45 D (±0.78). Measurement points M0-12 and differing Ns across those are explained in table 1.
(Enlarge Image)
Figure 3.
Absolute differences of JO (A) and J45 (B) between neighbouring measurement points (line within the box) with their respective quartiles (top and bottom of the box) and mean absolute differences (rhombus within the box, all ps<0.05). Repeated J0 measurements in the aphakic state revealed a mean difference of 0.31D (limits of agreement (d±1.96 s) +0.77 D and −0.8 D) and, respectively, of 0.41 –D (limits of agreement (d±1.96 s) +0.55D and −0.49D) in pseudophakia. Test-retest reliability amounted to 0.73 (aphakia) and 0.67 (pseudophakia). Repeated J45 measurements in the aphakic state revealed a mean difference of 0.36D (limits of agreement (d±1.96 s) +0.67 D and −0.71 D) and, respectively, of 0.43D (limits of agreement (d±1.96 s) +0.85D and −0.60D) in pseudophakia. Test-retest reliability amounted to 0.80 (aphakia) and 0.78 (pseudophakia). Measurement points M0-12 and differing Ns across those are explained in table 1.
Out of a gross number of 814 intraoperative measurement attempts, 462 readings succeeded in recording a wavefront map. Sixteen maps were excluded from analysis because the cylinder measurements were outside the dynamic range of the aberrometer (all spherical measurements were within range).
As expected, the WF map quality of excluded eyes was significantly lower than the one included (19.11% (SD 16.15) excluded versus 39.78% (SD 18.58) included eyes, p<0.001). ORs predicting measurement failure showed that only CDVA significantly influenced results at some instances as Table 2 shows.
Across all measurement points, all maps with refractions within the dynamic range of the aberrometer were graded in terms of their total area. The highest total number of successful readings (n=63) was achieved in aphakia with OVD. The highest (50.63%, SD 20.23) and lowest (29.19%, SD 13.94) quality of wavefront maps (WFM) were found after CCI and in pseudophakia with OVD, respectively (100% corresponds to the highest possible area; see figure 4).
(Enlarge Image)
Figure 4.
Mean wavefront map quality of consecutive intraoperative wavefront aberrometry (WA) measurements (with SD) in terms of relative WF map integrity (100% corresponds to the highest possible area). Measurement points M0-12 and differing Ns across those are explained in table 1.
The mean cyclotorsional deviation of axis for all eyes between the seated and the supine patient amounted to 3.84° (SD±64.8°, n=37). When comparing the last intraoperative supine reading with the upright one from the following day, a mean cyclotorsion of −14.5° (SD±68°, n=14) was noted. Both effects were statistically not significant (t(37)=0.365, p=0.359 and t(14)=−0.0.82, p=0.789).
Repeated-measure ANOVA revealed that the observed effects could not be explained by the surgeon factor. There was no significant interaction between either WF map quality across surgeons during repeated measurements in pseudophakia (F(4, 46)=2.189, p=0.085) or in aphakia (F(4, 38)=1.315, p=0.282). No statistically significant explanatory variables for threefold measurements in aphakia and pseudophakia were found (Table 3).
Table 4 describes a strong but not significant association between PNS grade and preoperative visual acuity in an expected positive relationship, that is, the higher the PNS grade, the higher the CDVA (LogMar).
No statistically significant relationship was found between PNS grade and WF quality scores (all ps>0.05).
Results
The study population encompassed 74 eyes (34 right and 40 left eyes) from 74 patients (42 male and 32 female). The mean age of the patients was 69.0 (SD 11.3) years, and the baseline manifest refraction was +0.31 sphere (SD 2.9) and −1.1 cylinder (SD 1.05).
Table 1 provides an overview of descriptive statistics on preoperative intraoperative and postoperative WA refractive data.
Figure 1A plots the median absolute differences of SE between neighbouring measurement points. The detailed differences between the preoperative measurement on the seated patient versus the first intraoperative measurement are shown in figure 1B. Repeated SE measurements with limits of agreements in aphakia and pseudophakia are shown in figure 2A and B. Test-retest reliability amounted to 0.98 and 0.88, respectively, which being >0.75 can be classified as good. Figure 3A and B plot the mean absolute differences of J0 and J45 between neighbouring measurement points (black circles).
(Enlarge Image)
Figure 1.
Absolute differences of spherical equivalent (SE). (A) Median absolute differences of SE between neighbouring measurement points (line within the box) with their respective quartiles (top and bottom of the box) and mean absolute differences (rhombus within the box, all ps<0.05). (B) Bland–Altman plot comparing preoperative SE measurements on the seated patient versus the first intraoperative measurement on the horizontal patient. Mean difference is shown by solid black line and limits of agreement by dashed lines. The 95% CIs for the lower and upper limits of agreement are −4.35 to −2.56 D and 3.62 to 1.83 D, respectively. Based on the sample size (N=38), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤ 0.9 D (±1.58). Measurement points M0-12 and differing Ns across those are explained in table 1.
(Enlarge Image)
Figure 2.
Spherical equivalent (SE) reproducibility in aphakia and pseudophakia; (A) Bland–Altman plot comparing threefold repeated SE measurements in aphakia. Mean difference is shown by solid black line and limits of agreement by dashed lines. The 95% CIs for the lower and upper limits of agreement are −0.96 to −0.47 D and 0.93 to 0.44 D, respectively. Based on the sample size (n=27), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤0.2 D (±0.36). (B) Bland–Altman plot comparing threefold repeated SE measurements in pseudophakia after removal of ophthalmic viscosurgical device (OVD). The 95% CIs for the lower and upper limits of agreement are −2.14 to −0.94 D and 2.13 to 0.93 D, respectively. Based on the sample size (n=22), a test power of 80% at the 0.05 significance level was achieved for a mean difference of ≤0.45 D (±0.78). Measurement points M0-12 and differing Ns across those are explained in table 1.
(Enlarge Image)
Figure 3.
Absolute differences of JO (A) and J45 (B) between neighbouring measurement points (line within the box) with their respective quartiles (top and bottom of the box) and mean absolute differences (rhombus within the box, all ps<0.05). Repeated J0 measurements in the aphakic state revealed a mean difference of 0.31D (limits of agreement (d±1.96 s) +0.77 D and −0.8 D) and, respectively, of 0.41 –D (limits of agreement (d±1.96 s) +0.55D and −0.49D) in pseudophakia. Test-retest reliability amounted to 0.73 (aphakia) and 0.67 (pseudophakia). Repeated J45 measurements in the aphakic state revealed a mean difference of 0.36D (limits of agreement (d±1.96 s) +0.67 D and −0.71 D) and, respectively, of 0.43D (limits of agreement (d±1.96 s) +0.85D and −0.60D) in pseudophakia. Test-retest reliability amounted to 0.80 (aphakia) and 0.78 (pseudophakia). Measurement points M0-12 and differing Ns across those are explained in table 1.
Out of a gross number of 814 intraoperative measurement attempts, 462 readings succeeded in recording a wavefront map. Sixteen maps were excluded from analysis because the cylinder measurements were outside the dynamic range of the aberrometer (all spherical measurements were within range).
As expected, the WF map quality of excluded eyes was significantly lower than the one included (19.11% (SD 16.15) excluded versus 39.78% (SD 18.58) included eyes, p<0.001). ORs predicting measurement failure showed that only CDVA significantly influenced results at some instances as Table 2 shows.
Across all measurement points, all maps with refractions within the dynamic range of the aberrometer were graded in terms of their total area. The highest total number of successful readings (n=63) was achieved in aphakia with OVD. The highest (50.63%, SD 20.23) and lowest (29.19%, SD 13.94) quality of wavefront maps (WFM) were found after CCI and in pseudophakia with OVD, respectively (100% corresponds to the highest possible area; see figure 4).
(Enlarge Image)
Figure 4.
Mean wavefront map quality of consecutive intraoperative wavefront aberrometry (WA) measurements (with SD) in terms of relative WF map integrity (100% corresponds to the highest possible area). Measurement points M0-12 and differing Ns across those are explained in table 1.
The mean cyclotorsional deviation of axis for all eyes between the seated and the supine patient amounted to 3.84° (SD±64.8°, n=37). When comparing the last intraoperative supine reading with the upright one from the following day, a mean cyclotorsion of −14.5° (SD±68°, n=14) was noted. Both effects were statistically not significant (t(37)=0.365, p=0.359 and t(14)=−0.0.82, p=0.789).
Repeated-measure ANOVA revealed that the observed effects could not be explained by the surgeon factor. There was no significant interaction between either WF map quality across surgeons during repeated measurements in pseudophakia (F(4, 46)=2.189, p=0.085) or in aphakia (F(4, 38)=1.315, p=0.282). No statistically significant explanatory variables for threefold measurements in aphakia and pseudophakia were found (Table 3).
Table 4 describes a strong but not significant association between PNS grade and preoperative visual acuity in an expected positive relationship, that is, the higher the PNS grade, the higher the CDVA (LogMar).
No statistically significant relationship was found between PNS grade and WF quality scores (all ps>0.05).
Source...