ESTIMASI JUMLAH SAMPEL UNTUK SEROSURVEY PASCAVAKSINASI PMK 2023_p2h_yy27feb23_final.pptx

PMK

1. DESAIN SURVEY LAPANG MONITORING KEKEBALAN TERHADAP PMK TAHUN 2023 KELOMPOK SUBSTANSI P2H-DITKESWAN KEMENTERIAN PERTANIAN REPUBLIK INDONESIA Hotel Santika -Depok, 20-21 Februari 2023

2. 2 Metodologi : Estimation of immune proportions using a simple random sampling (SRS) selection process: To estimate the proportion of animals vaccinated for the first time that will develop a specific level of antibodies considered protective and due to the vaccine. Estimation of immune proportions using a more complex study design (two-stage random sampling): Objective : To estimate the proportion of animals with a ‘detectable level of antibodies’ in the population.

3. 1. Estimation of immune proportions using a simple random sampling (SRS) selection process Objective : To estimate the proportion of animals vaccinated for the first time that will develop a specific level of antibodies considered protective and due to the vaccine. Target population : The 6- to 12-month-old animals that will be injected with FMD vaccine not previously exposed to natural infection . Measurable response : SP antibody titres against the types of virus contained in the vaccine and NSP antibodies . If a threshold of protection is known , then animals with SP antibody titres above such a threshold are considered ‘adequately protected’ and those animals with titres below are considered ‘not adequately protected.

4. Methodological approach and implications : As the objective is to estimate the immune response due to the injection of the vaccine , differentiation between antibodies due to the vaccine and antibodies due to previous exposure to field virus is necessary : Assuming that the vaccine used does not induce detectable NSP antibodies , testing serum samples also for NSP antibodies in each interval allows a distinction to be made between antibodies due to vaccination and antibodies due to infection. When incidence of virus circulation in the area is either very low or zero, tests for NSP may be used to indicate NSP purity (actually specificity of NSP tests in vaccinated animals). Sampling animals at different intervals, as proposed, implies that the sampling plan must be prepared in advance and those animals should be individually identified (i.e. ear-tagged).

5. Waktu sampling Time of sampling : This sub-category should be sampled at the time of vaccination ( t 0) and after 28, 56 and 168 days ( t 1, t 2, t 3, respectively). This allows an assessment of the immune response induced by the vaccine, and the duration during the campaign , and provides a way of discounting previous exposure to the virus or virus circulation during the field trial.

6. Desain study A n individual list of animals in the target population is usually not available in advance (especially in developing countries) and thus it may be impossible to strictly adhere to an SRS procedure. A practical approach to overcome this issue is to make a preliminary selection of 10 to 15 epi-units before the vaccination campaign starts (the number of 10–15 is only indicative and in general the number of epi-units should be sufficient to yield a number of eligible animals at least twice the estimated sample size). The selection of the epi-units should be based on knowledge of the past occurrence of FMD (to ensure that sampled animals are less likely to have been already exposed to field virus). Once the epi-units have been selected, they should be visited to create a census of all animals age matching the eligibility criteria (source population from which individuals will be selected). If animals could be individually ear-tagged at the time of the visit , a list could be created from which animals could be subsequently selected under an SRS procedure (a systematic random sampling approach could also be applied). This approach, from the practical point of view, can be considered a proxy for an SRS design.

7. SAMPLE SIZE Estimation of the sample size involves both non-statistical and statistical considerations. Non-statistical considerations include availability of sampling frames, resources, manpower and facilities. According to the criteria indicated in section 3.4: ( i ) the expected proportion is 85%, indicated as p in equation 1); (ii) the absolute error (allowable error or desired precision, indicated as e in equation 1) is 10% (which means that if the expected proportion is really 85%, it is expected that the estimate obtained will lie between 75% and 95%; and (iii) finally the level of confidence chosen is 95% (which means that the investigator wants to be 95% confident that the estimate of the proportion (should that be truly 85%) will actually lie between 75% and 95%. The value 1.96 is the normal standard deviation for a 95% confidence level (should the investigator wish to have a 99% or 90% confidence level, the value 1.96 should be replaced with 2.58 or 1.64, respectively).

8. 2. Estimation of immune proportions using a more complex study design ( two-stage random sampling ) Objective : To estimate the proportion of animals with a‘detectable level of antibodies’ in the population Target population : The total number of animals present in the area or zone where the vaccination programme is implemented and for which the conclusions of the survey will apply. Outcome of interest : Level of detectable antibodies against FMD. Measurable response : SP antibody titre against the types of virus contained in the vaccine. If a threshold of protection is known, then animals with antibody titres above such a threshold can be considered ‘adequately protected’ and those animals with titres below can be considered ‘not adequately protected.

9. WAKTU SAMPLING When a vaccination programme is regularly implemented, immunity can be estimated either at the time that the highest or lowest level is expected. This is either 30 days post vaccination or the day animals will be vaccinated again, respectively. If vaccination is not regularly implemented, sampling on the day of vaccination may not be relevant.

10. Desain study The assessment of the level of immunity of the general population or of specific sub-groups generally involves the design of a complex survey . In this specific case, it is likely that the design will be a two-stage cluster sampling , with the first stage being the epi-units (primary sampling units – PSUs) and the second stage individual animals (secondary sampling units – SSUs) present within those selected PSUs. Obviously, this procedure restricts the choice of the SSUs to only those PSUs selected in the first stage.

11. Penghitungan Besaran Sampel per Provinsi Catatan : Digunakan untuk tujuan monitoring Pasca Vaksinasi PMK pada suatu lokasi ( baik yang sudah dan belum divaksinasi ) untuk mengetahui Herd immunity = kekebalan kelompok Untuk tujuan lain, mengacu pada SE Pengamatan PMK sebelumnya .

12. P=0.70 (adjusted from Foot and mouth disease vaccination and post-vaccination monitoring. Guidelines) e=0.05, 0.07. 0.1 N= jumlah sapi dan kerbau per provinsi rho= 0.3 b= sampel per cluster= 20 ekor I. n.crude = epi.sssimpleestb (N , P= 0.70, epsilon = 0.05, error = "relative", se = 1(?), sp = 1(?), nfractional = FALSE, conf.level = 0.95) II. Design effect: D =0.3 * (20 – 1) + 1; D= 6.7 III. Besaran sampel adjusted n.adj =7.6 * n.crude IV. Jumlah unit/cluster/ psu PSU= n.adj /b EpiR , asumsi “imperfect” diagnostic test Penghitungan Besaran Sampel

13. Besaran sampel per provinsi Jumlah cluster/ psu Besaran sampel adjusted Besaran sampel

14. Kabupaten dan cluster/ psu /unit epidemiologi ditentukan oleh masing-masing BV/BBV Besaran sampel dan cluster adjusted sudah ditentukan Pilih kabupaten berbasis risiko dan kemungkinan akses ke populasi Pilih Desa /unit berbasis risiko dan kemungkinan akses ke populasi

15. TERIMA KASIH