Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
ABSTRACT
Purpose: To assess and compare the accuracy of 2 methods of predicting visual acuity
after phacoemulsification.
Results: Both the interferometer and PAM underestimated the retinal visual capacity.
Three-degree white-light interferometry gave significantly better mean predicted results
than 8-degree interferometry and the PAM. There was no statistically significant dispar-
ity between predicted and postoperative results with 3-degree interferometry (1.04 ⫾
0.57 logMAR; – 0.09 ⫾ 0.27 decimal) (P ⫽ .0647) and a statistically significant disparity
with 8-degree interferometry (0.89 ⫾ 0.59 logMAR; – 0.13 ⫾ 0.27 decimal) and the PAM
(0.66 ⫾ 0.62 logMAR; – 0.22 ⫾ 0.24 decimal) (P ⫽ .0001). The predicted values were
widely dispersed; the correlation indices were 0.38 with the PAM (P ⫽ .091), 0.39 with
3-degree interferometry (P ⫽ .001), and 0.49 with 8-degree interferometry (P ⫽ .0005).
1 75 M 0.40 (0.40) 0.80 (0.10) 0.67 (0.17) 0.80 (0.10) 0.90 (0.05)
2 64 F 0.40 (0.40) NA 1.00 (0.00) 0.50 (0.30) 0.80 (0.10)
3 64 F 0.20 (0.70) 0.33 (0.48) 0.33 (0.48) 0.40 (0.40) 0.90 (0.05)
4 80 F 0.40 (0.40) 1.00 (0.00) 0.67 (0.17) 0.50 (0.30) 0.90 (0.05)
5 70 F 0.40 (0.40) 0.67 (0.17) 0.67 (0.17) 0.50 (0.30) 0.60 (0.22)
6 70 F 0.40 (0.40) 0.50 (0.30) 0.67 (0.17) 0.80 (0.10) 0.60 (0.22)
7 80 F 0.30 (0.52) 0.80 (0.10) 0.67 (0.17) 0.33 (0.48) 0.90 (0.05)
8 63 F 0.40 (0.40) 1.00 (0.00) 1.00 (0.00) 0.40 (0.40) 0.80 (0.10)
9 92 F 0.10 (1.00) NA 0.05 (1.30) 0.50 (0.30) 0.10 (1.00)
10 61 F 0.20 (0.70) 0.80 (0.10) 0.33 (0.48) 0.25 (0.60) 0.30 (0.52)
11 69 F 0.40 (0.40) 0.80 (0.10) 0.67 (0.17) 0.80 (0.10) 0.80 (0.10)
12 28 M 0.40 (0.40) 1.00 (0.00) 1.00 (0.00) 0.80 (0.10) 1.00 (0.00)
13 68 M 0.10 (1.00) 0.02 (1.60) 0.02 (1.60) 0.10 (1.00) 0.10 (1.00)
14 88 M 0.40 (0.40) 1.00 (0.00) 0.80 (0.10) 0.50 (0.30) 0.80 (0.10)
15 82 F 0.20 (0.70) NA 0.05 (1.30) 0.10 (1.00) 0.30 (0.52)
16 78 F 0.40 (0.40) 1.00 (0.00) 1.00 (0.00) 0.40 (0.40) 0.50 (0.30)
17 64 F 0.30 (0.52) 0.67 (0.17) 0.40 (0.40) 0.50 (0.30) 0.70 (0.15)
18 77 F 0.20 (0.70) 0.50 (0.30) 0.50 (0.30) 0.28 (0.55) 0.50 (0.30)
19 78 F 0.40 (0.40) 0.80 (0.10) 0.40 (0.40) 0.50 (0.30) 0.70 (0.15)
20 79 F 0.30 (0.52) 0.67 (0.17) 0.67 (0.17) 0.33 (0.48) 0.90 (0.05)
21 74 F 0.40 (0.40) 0.80 (0.10) 1.00 (0.00) 0.66 (0.18) 0.90 (0.05)
22 77 M 0.50 (0.30) 0.80 (0.10) 0.67 (0.17) 0.80 (0.10) 0.90 (0.05)
23 30 M 0.05 (1.30) 0.02 (1.60) 0.02 (1.60) 0.50 (0.30) 0.80 (0.10)
24 57 M 0.40 (0.40) 0.50 (0.30) 0.50 (0.30) 0.33 (0.48) 1.00 (0.00)
25 74 M 0.40 (0.40) 0.80 (0.10) 0.80 (0.10) 0.50 (0.30) 1.00 (0.00)
26 85 M 0.40 (0.40) 0.80 (0.10) 0.80 (0.10) 0.66 (0.18) 0.80 (0.10)
27 76 F 0.20 (0.70) 0.50 (0.30) 0.50 (0.30) 0.40 (0.40) 1.00 (0.00)
28 78 F 0.30 (0.52) 0.02 (1.60) 0.33 (0.40) 0.50 (0.30) 0.50 (0.30)
29 79 F 0.40 (0.40) 0.33 (0.40) 0.20 (0.70) 0.50 (0.30) 0.80 (0.10)
30 85 M 0.40 (0.40) 1.00 (0.00) 0.80 (0.10) 0.80 (0.10) 1.00 (0.00)
31 62 F 0.16 (0.80) 0.25 (0.60) 0.30 (0.52) 1.00 (0.00) 0.70 (0.15)
32 73 F 0.40 (0.40) 1.00 (0.00) 0.80 (0.10) 0.66 (0.18) 1.00 (0.00)
33 75 M 0.10 (1.00) 0.33 (0.40) 0.25 (0.60) 0.28 (0.55) 0.60 (0.22)
34 76 M 0.40 (0.40) 0.67 (0.17) 0.67 (0.17) 0.50 (0.30) 0.70 (0.15)
35 80 F LP NA NA NA LP
36 80 F 0.16 (0.80) 0.67 (0.17) 0.67 (0.17) 0.50 (0.30) 1.00 (0.00)
37 62 M 0.40 (0.40) 0.80 (0.10) 0.67 (0.17) 0.66 (0.18) 0.60 (0.22)
38 73 F 0.40 (0.40) 1.00 (0.00) 0.80 (0.10) 0.80 (0.10) 0.80 (0.10)
39 78 F 0.10 (1.00) 0.67 (0.17) 0.40 (0.40) 0.20 (0.70) 1.00 (0.00)
40 23 F 0.05 (1.30) 0.40 (0.40) 0.80 (0.10) NA 0.10 (1.00)
Table 1. (cont.)
41 74 F 0.30 (0.52) 0.67 (0.17) 0.50 (0.30) 1.00 (0.00) 0.90 (0.05)
42 76 F 0.40 (0.40) 1.00 (0.00) 1.00 (0.00) 1.00 (0.00) 0.80 (0.10)
43 19 M 0.05 (1.30) 0.50 (0.30) 0.40 (0.40) 0.28 (0.55) 0.50 (0.30)
44 81 M 0.30 (0.52) 0.67 (0.17) 0.67 (0.17) 0.50 (0.30) 0.60 (0.22)
45 78 F 0.40 (0.40) 0.20 (0.70) 0.40 (0.40) 0.66 (0.18) 0.50 (0.30)
46 80 F 0.40 (0.40) 0.80 (0.10) 0.67 (0.17) 1.00 (0.00) 0.90 (0.05)
47 73 M 0.40 (0.40) 1.00 (0.00) 1.00 (0.00) 0.80 (0.10) 0.90 (0.05)
VA ⫽ visual acuity; IF-3 ⫽ 3-degree interferometry; IF-8 ⫽ 8-degree interferometry; PAM ⫽ potential acuity meter; LP ⫽ light perception; NA ⫽
not applicable
ferometry and the PAM (– 0.58 to 0.30 logMAR) (P ⫽ sion test showed that individual prediction may be
.0216), or 8-degree interferometry and the PAM (P ⫽ overestimated or underestimated from ⫹0.25 to – 0.15
.003) (Figure 2). Therefore, even though 3-degree inter- lines on the decimal scale around the mean result. This
ferometry appeared to be the best predictor, the disper- was also true with 8-degree interferometry (⫹0.31 to
– 0.16 lines, decimal) and the PAM (⫹0.50 to – 0.26
lines, decimal). The PAM had a wider dispersion of
results than either interferometry method when individ-
ually results were considered; the difference was not sta-
tistically significant, however.
There was a significantly poor individual correla-
tion between predicted and postoperative visual acuities
with all 3 methods: 3-degree interferometry (r ⫽ 0.39,
P ⫽ .01); 8-degree interferometry (r ⫽ 0.49, P ⫽
.0005); PAM (r ⫽ 0.38, P ⫽ .0091).
Discussion
In this comparative study, 3-degree white-light in-
terferometry was more reliable in predicting postopera-
tive visual acuity than 8-degree interferometry or the
PAM; the mean prediction with 3-degree white-light
interferometry was approximately 1 line (decimal scale)
from the final visual acuity. However, all 3 methods
underestimated postoperative visual acuity, which may
have relatively little effect on patient satisfaction. It may
be of concern to surgeons evaluating the risk/benefit
ratio of cataract surgery. All 3 methods had a tendency
toward poor prediction of individual acuity as the dis-
persion of results was significantly wide and the correla-
tion indices unsatisfactory. Thus, none of the methods
can predict final visual acuity in a linear mode and a
prediction of poor visual recovery may not reflect an
individual’s visual capacity.
All the patients in our study had mild to moderate
Figure 2. (Le Sage) Dispersion of predicted versus final visual
acuity. Top: 3-degree interferometry (IF-3°). Middle: 8-degree inter- cataract. There were no cases of mature cataract, which
ferometry (IF-8°). Bottom: PAM (x- and y-axis ⫽ logMAR scale). is more likely to induce a false negative (underestimated)
predicted visual acuity. Other conditions reported to interferometry. This in agreement with the results of
induce inaccurate prediction of visual recovery are pos- Goldmann and coauthors,20 who suggest that small
terior subcapsular cataract, high myopia, ARMD, cys- fields are more accurately centered on the fovea and may
toid macular edema, retinal detachment, amblyopia, be more accurate in patients suspected of having macu-
nystagmus, glaucoma, and miosis.17–23 lar disease. This prevents test recognition from the
In our study, 4 of the 5 myopic patients displayed a parafoveal retina.
strong disparity in predicted visual acuity and final vi- In our study, a single trained examiner performed all
sual acuity between the 2 interferometry tests and be- visual acuity prediction tests after providing patients a
tween the interferometer and the PAM. The orientation detailed explanation. This increased the reliability of the
of the light rays reaching adjacent areas of the retina may results. Miller et al.29 propose that the poor correlation
be a cause of this disparity.19,20 Thus, the 3 methods indices in their series of 82 patients obtained with both
seemed inaccurate in predicting final acuity in eyes with interferometry (r ⫽ 0.03) and the PAM (r ⫽ 0.07) were
high myopia. This is in contrast to the results of Datiles the result of the technician’s lack of experience as well as
et al.,24 who report accurate predicted acuity in 3 cases the patients’ ages and associated diseases, which lead to
of high myopia. Datiles et al. stress that the low preop- poor cooperation. These biases result in a poor estima-
erative visual acuity in high myopia may be the result of tion of visual acuity recovery and thus contradictory
amplified distortion when the light passes through even outcomes among studies.25,28 –31
a moderate cataract. Few studies have compared the use of interferom-
The predicted acuity was underestimated by inter- etry and the PAM in ECCE. Most did not provide
ferometry in all 4 patients with posterior subcapsular a statistical analysis to determine whether interferom-
cataract and by the PAM in 3 of the patients. The results etry or the PAM is significantly accurate. Spurny and
support the findings of Lasa and coauthors21 and coauthors25 compared a white-light interferometer
Schraub and coauthors,23 who used a Rodenstock威 reti- (Lotmar visometer, Haag-Streit) with the PAM in
nometer and the PAM to predict acuity. Both groups 54 eyes having cataract surgery or neodymium:YAG
hypothesize that this underestimation is linked to the laser capsulotomy. Predictions were more accurate
increased central density of the opacity in a nuclear cat- with the visometer than the PAM, supporting our
aract; the opacity is more diffuse and regular, giving results. Strong30 compared the Rodenstock laser
more reliable values. interferometer, Lotmar visometer, and white-light
As previously described,17,18,21,25 interferometry IRAS interferometer in 14 eyes. The laser interferome-
accurately predicted visual acuity in patients with atro- ter was the most accurate, but no statistical analysis
phic ARMD and overestimated it in the patient with was reported. Graney et al.28 found laser interferometry
exudative ARMD. The PAM results were more accurate and the PAM less accurate than a clinical index based on
in the case of exudative ARMD. age, preoperative visual acuity, number of current pre-
The postoperative visual acuity matched the pre- scription medications, and the ability to read a newspa-
dicted outcome with all 3 methods in patients with pig- per. Nevertheless, these studies had several biases as
ment epithelium retinopathy, which agrees with the different ophthalmologists performed the prediction
findings in a previous study.26 The results in the patient tests and all patients were from middle- or upper-class
with amblyopia were also accurate, which is in contrast families. However, the PAM was statistically more accu-
to previous studies showing overestimation in such cas- rate in predicting acuity than laser interferometry. In a
es.9 In the patient with nystagmus, interferometry over- prospective study of 35 eyes, Datiles et al.24 found that
estimated and the PAM underestimated final visual the PAM and laser interferometer were equally good
acuity, as reported in a previous study.27 However, con- predictors in cases of mild to moderate cataract. Gold-
clusions cannot be made based on the evaluation of a stein et al.31 found better results with the Lotmar vi-
single individual. someter than with the IRAS interferometer. Therefore,
There was a significant disparity among the 3 meth- it appears that these studies had several strong examiner
ods in predicting visual acuity recovery after phacoemul- and patient biases, which may account for the contra-
sification. The best results were obtained with 3-degree dictory results.
Meter using a minute aerial pinhole aperture. Ophthal- 30. Strong N. Interferometer assessment of potential visual
mology 1983; 90:1360 –1368 acuity before YAG capsulotomy: relative performance of
28. Graney MJ, Applegate WB, Miller ST, et al. A clinical three instruments. Graefes Arch Clin Exp Ophthalmol
index for predicting visual acuity after cataract surgery. 1992; 230:42– 46
Am J Ophthalmol 1988; 105:460 – 465 31. Goldstein J, Jamara RJ, Hecht SD, et al. Clinical com-
29. Miller ST, Graney MJ, Elam JT, et al. Predictions of parison of the SITE IRAS hand held interferometer and
outcomes from cataract surgery in elderly persons. Oph- Haag-Streit Lotmar visometer. J Cataract Refract Surg
thalmology 1988; 95:1125–1129 1988; 14:208 –211