• No se han encontrado resultados

IV. RESULTADOS Y DISCUSION

4.3. Estado sanitario de la infraestructura de abastecimiento

Averaae Motion Seen Motion Sensitivity

area SE 95% C. I. area SE 95% C.l.

Group 1 0.66 0.04 0.57, 0.75 0.7 0.05 0.61, 0.76

Group 2 0.75 0.06 0.64, 0.86 0.9 0.04 0.8, 0.95

Group 3 0.94 0.02 0.90, 0.98 1.0 0.02 0.94, 1.00

* Power for testing one side test of significance with p = 0 .05 (Colton, 1974)

There w ere no significant differences between A M S and MS for detecting the sarne group patients in terms of AUC, despite the fact that M S seems to provide more potential cutoffs to improve the sensitivity while the specificity is not seriously reduced. For example, in Group 3, a cut-off point of 8 /1 0 to 1 0 /1 0 would give a sensitivity of 8 0 % for A M S and 9 5 % for M S, respectively. For the same cut off point, specificity of 6 5 % was obtained for A M S compared

to a specificity of 3 0 % for MS. The sensitivities and the specificities corresponding to the optimized cut-off point are given in Table 3 -1 4 .

Table 3-14 Sensitivity and specificity with the optimized cut-off

Averaoe Motion Seen Motion Sensitivity GHT

Group cutoff Sensitivity Specificity cutoff Sensitivity Specif

67% 92% 94% Group 1 9.0 39% 92% 8/10 51% Group 2 9.0 56% 92% 7/10 67% Group 3 8 . 0 85% 98% 6/10 92%

3 -5 -2 Application in screening for ocular diseases in Nigeria

This part of the study was repeated one year later after the initial examination. Because the data from Second visit was incomplete at the end of 1 9 9 2 , the results presented here were mainly based on the initial examination data. There w ere tw o population samples tested by M S S T. One was in Kaduna city which has been considered to be a non-endemic area for onchocercal optic nerve disease. The second population was in the meso-endemic onchocercal communities which consisted of isolated and illiterate rural village people.

2 3 1 4 M SST files were originally recorded: 1 2 3 8 from the W HO project and 1 0 7 6 from others. Except 4 5 4 individuals, all were from the endemic area. Because the data from outside the W H O project did not include a complete eye examination, they w ere only used for assessment of risk of motion loss based upon a population rather than an individualisée section 1-2-1). In the W HO project, forty-four files were excluded because of unreliable data, such as missing identification, age and sex in their data files. Of the remaining total, there were 3 7 5 individuals from the WHO survey.

Sim plicity & A cceptance

Chapter 3 Results 1 1 8

There were ten village helpers and 11 ophthalmic nurses trained to perform M SST. Among village helpers, one was a trainee from a church, 8 were primary school graduates and one had no school background. The training time was approximately 1 hour for village helpers. All ophthalmic nurses were students attending an ophthalmic nurse training course at the ABU Teaching Hospital, except 2 who were members of the W HO project. All had a half hour training for M SST in the hospital. Except for one village helper, all of them could conduct the test satisfactorily. This allowed 6 8 % of M SST to be done by local people or paramedics (Table 2 -7 ). Since M SST is less affected by the m ulticontrast environment, no special effort was made to have a dark room for the test. Public rooms, such as school rooms and the waiting room in the hospital and huts in each village were used. The vast majority of the subjects understood w hat they should do during the test after they had seen other people tested (see below). Few tests(2.1% ) were incomplete due to lack of understanding. The rural communities had no objection to home visits.

In the hospital, M SST was used for 6 days. A large number of patients or patients relatives or their friends from the out-patient department were waiting to have M SST examination. There were always tw o queues in front of each computer. In addition, there was a 10 0 % response rate from 160 individuals who were required to be retested with M SST on more than tw o different occasions. No one complained that the test was uncomfortable. Rather, the test attracted many people even though they did not know w h at the test was for. Although there was no advertisement for MSST, many people wanted to have M SST because they had heard about it from friends, colleagues and children.

Rapid screening te st

To start MSST, it is necessary only to switch on the computer. A batch file manages MSST and generates the stimuli for the test. The testing time per eye for 5 repeats and 10 repeats w ere 9 8 seconds and 167 seconds, respectively. The testing time for 5 repeats for both eyes including training was 5 .5 minutes. For most subjects additional training was not necessary because they quickly became confident to do the test from observing other people doing it. The maximum number of visual fields screened in one day was 2 2 5 w ith tw o Sharp

computers. It was always possible to retest immediately if this was required(e.g. for intra­ subject variation).

M otion sensitivity vs Microfilaria! Load

When the cut off point of M SST was 6 /1 0 , the sensitivity of detecting OND was 9 1 % in 56 OND cases defined by the WHO project and the specificity was 7 5 % in 3 1 9 Nigerians without OND. When 2 5 cases were excluded: 9 with low vision (VA < 6 /1 8 ), 7 with early cataract and 9 potential glaucoma cases (cup/disc ratio > 0 .5 and/or intra ocular pressure > 21 mmhg), the specificity was still low er(82% ) than was found in abnormal motion case finding in hospital.

Table 3-15 D is tr ib u tio n o f m ic r o f ila r ia l load by c l i n i c a l status 'N orm al'N igerian*

n= 255 MSST Neg. Pos. OND n=56 MSST Neg. Pos Total n=311 MSST Neg. Pos. M ic r o f ila r ia l Load (mf/mg) < 10 (n=224) 10.1-30 (n= 46) >30 (n= 41) n=201 n=54 188 21 8 17 5 16 n=5 n=51 4 11 1 20 20 n=206 n=105 192 32 9 37 5 36 p value 0.0000 0.0000

* 25 cases with other ocular abnorm alities and 39 Nigerians who had no m ic r o f ila r ia l load record were excluded

The relationship between Motion sensitivity versus Microfilarial load was therefore studied. Table 3 -1 5 shows the distribution of MSST by clinical group according to their microfilarial load. Among subjects with normal ocular examination and with a microfilarial load greater than 10 m f/m g, the prevalence of abnormal MSST was roughly 7 times higher than subjects who had a microfilarial load less 10 mf/mg(p < 0 .0 0 0 ). If these cases a the microfilarial load greater than 10 mf/mg were excluded, the specificity became 9 0 % (1 8 8 /2 0 9 ).

Observer agreement

Intra-observer agreement on MSST was measured by repeating M SST immediately at the same sitting. Inter-observer agreement for each subject on M SST was separately measured on tw o different MSST sittings using different computers and different operators in the same testing room after 5 minutes break. Each of the 1 12 eyes(71 individuals) was examined twice by each of three operators for inter-observer variation.

Chapter 3 Results 1 2 0

Documento similar