Sample Size. As this is an extension study, no formal sample size analysis was performed for this study. Study data will be tabulated and summarized.
Sample Size. For each of the locations, the sample size was determined at 5% level of precision (also called desired margin of error), 95% confidence level and 50% population proportion of response to key study indicators. The choice of population response proportion is arbitrary; but this is what is assumed to generate the largest possible sample in the absence of a prior knowledge about population response to key study indicators. The sample size was not adjusted for non- response factor. Like in many surveys, non-response of the sampling unit (here household) has been tackled by substituting original sample unit by another.
Sample Size. 10.5.1 The number of passengers interviewed during each Fare Evasion Survey will be sufficient to ensure that the overall Fare Evasion Rate for bus services, calculated from each Fare Evasion Survey, shall have an overall sample error of less than 1%.
Sample Size. The sample size depended upon when theoretical saturation is reached and was estimated to range from 20 to 30 participants for service providers group and 20 to 30 for ‘people with mental illness and their primary caregivers’ group’ This included both men and women. Where possible, equal numbers of men and women were recruited with participants representing all adult age ranges (age 18 years and above). A total of 32 in-depth interviews were conducted: 10 with people with mental illness, 6 with caregivers and 16 with service providers. We conducted one focus group discussion (FGD) with service users and their caregivers. There were sixteen participants in the FGD. The participants in the FGD included six service users and ten caregivers.
Sample Size. To select the sample we will consider the 88 students of the Eight Grade of Basic General Education, equivalent to 11% of the students of the population's total in study, 2 English teachers which is 10% of the English teacher’s population.
Sample Size. The primary outcome for the interrater agreement analysis was the intraclass correlation coefficient (ICC) of the AHI. Given nine sleep scorers (one at each site), the 15 PSGs had a power of 83% to detect an ICC of at least 0.90, assuming a null hypothesis of ICC = 0.70. Statistical Analysis The interrater reliability measures used to examine the agreement among the nine different scorers were the ICC for continuous variables (respiratory indices, duration of sleep stages, and arousal index) and the kappa (κ) statistic for mul- tiple raters for the categorical variables (sleep stages). The levels of agreement using the ICCs of respiratory indices, dura- tion of sleep stages, and arousal index were classified as fol- lows: 0.00-0.25 = little, 0.26-0.49 = low, 0.50-0.69 = moderate, Table 1—Interscorer reliability of respiratory events Sleep Center (mean) AHI (events/h) 25.8 ± 19.4 22.7 ± 18.4 25.7 ± 19.0 25.9 ± 19.8 24.3 ± 17.2 22.7 ± 17.4 23.0 ± 15.4 25.9 ± 21.3 23.6 ± 18.1 0.95 0.91-0.98 Total Apneas 54.1 ± 71.9 31.6 ± 68.2 45.3 ± 66.5 46.3 ± 63.9 42.3 ± 60.3 20.3 ± 41.8 58.1 ± 78.3 104.6 ± 127.7 51.1 ± 72.4 0.73 0.55-0.88 Obstructive 44.1 ± 61.3 28.7 ± 68.2 39.5 ± 67.1 35.9 ± 55.2 37.5 ± 57.1 17.7 ± 36.7 56.4 ± 76.6 100.7 ± 127.7 42.2 ± 68.9 0.70 0.51-0.86 Central 6.7 ± 9.1 2.9 ± 7.1 4.4 ± 7.8 7.3 ± 10.2 1.0 ± 2.0 0.9 ± 1.4 0.9 ± 1.9 3.3 ± 6.2 6.7 ± 7.8 0.46 0.27-0.70 Mixed 3.3 ± 9.5 0.1 ± 0.3 1.5 ± 3.0 3.1 ± 7.5 4.0 ± 7.5 1.7 ± 5.4 0.8 ± 2.3 0.6 ± 1.2 2.2 ± 5.6 0.42 0.24-0.67 Hypopneas 90.1 ± 76.6 100.7 ± 85.7 109.1 ± 85.4 94.8 ± 73.8 108.2 ± 79.4 109.2 ± 87.4 81.1 ± 58.1 51.0 ± 36.9 86.5 ± 75.1 0.80 0.65-0.91 ODI (events/h) 22.4 ± 17.6 22.0 ± 17.5 24.6 ± 18.3 21.6 ± 17.0 25.3 ± 17.5 22.4 ± 18.0 19.7 ± 14.2 22.9 ± 17.4 20.9 ± 16.6 0.97 0.93-0.99 AHI, apnea-hypopnea index; CI, confidence interval; ICC, intraclass correlation coefficient; ODI, oxygen desaturation index. Downloaded from xxxxx://xxxxxxxx.xxx.xxx/sleep/article/36/4/591/2595972 by guest on 02 October 2020 Table 2—Kappa (κ) statistics evaluating epoch-by epoch sleep staging (NREM combined) W NREM REM All stages κ 0.78 0.77 0.78 0.78 95% CI 0.77-0.79 0.76-0.78 0.77-0.79 0.77-0.78 CI, confidence interval; NREM, nonrapid eye movement; REM, rapid eye movement. Number of epochs Table 3—Kappa (κ) statistics evaluating epoch-by-epoch staging (NREM separated) W N1 N2 N3 R All stages κ 0.78 0.31 0.60 0.67 0.78 0.63 95% CI 0.77-0.79 0.30-0.32 0.59-0.61 0.65-0.69 0.77-0.79 0.62-0.63 CI, confidence interva...
Sample Size. The original design of the study included 36 In-Depth Interviews (IDIs), However, a total of 44 interviews were conducted as new themes continued to emerge during data collection, and code saturation (Hennink, 2017) had not yet been reached. IDIs were divided equally between women and men, with participants and their husbands in the intervention arm accounting for 24 of the 44 total interviews. A total of twelve FGDs were conducted, divided equally between women and men in intervention and control arms of the study. The number of FGDs was based on recommendations to conduct at least two focus groups per stratum to identify meaning saturation (Hennink, 2019), with the strata in this study being gender and study arm. Details on the sampling structure are given in Table 1 below.
Sample Size. Sample size estimation has been done for the pivotal cohort of the study, i.e. patients receiving NI-0501 in second line. A minimum of 28 patients treated with NI-0501 in second line will be enrolled in the study. Sample size calculation is based on the primary efficacy endpoint of “Overall Response Rate”. Assuming an Overall Response Rate of 70%, the study will have 90% power to show a significant improvement above 40% using an exact binomial test33 at a one-sided significance level of 2.5%. Due to the rarity of primary HLH, the recruitment is competitive across all EU and US sites in order to gather data in a reasonable timeframe.
Sample Size. Incidence cohort study The study sample size is based on a generalized estimating equations (GEE) modeling approach that takes into account the correlated nature of the longitudinal measurements within study children. A Poisson distribution will be used to model the incidence of malaria in study children over the 12-month study period, with the aim of isolating the effect of recent LLIN use, controlling for other important confounding variables and predictors. To calculate the sample size needed, we used a program developed based upon a GEE approach to longitudinal data analysis [14]. Assumptions made for key parameter inputs into the model are summarized in Table 2. Table 2: Key parameters used in sample size calculation[14] Parameter Value used in calculation Power 70% Alpha (Type 1 error rate) 5% Malaria incidence among non-LLIN users 1 episode/child/year Malaria incidence among LLIN users 0.7 episodes/child/year LLIN use in study area among cohort children 70% Phi (within-child correlation) 0.25 Type of GEE correlation structure Exchangeable Dispersion parameter (Psi) 1 (model not over- or under-dispersed) Inputting the parameters specified above in the sample size program yields a total of 684 LLIN users and 297 non-LLIN users (total = 981 cohort study children) required to demonstrate at least a 30% reduction in incidence between the two groups. Assuming a 5% refusal rate, a 15% loss-to-follow-up rate across the study, and a mortality rate among the cohort of 65 per 1,000 live births (~1.5% per year), a total of 1,335 children need to be enrolled at baseline. Under the assumption that in the study area, 60% of households have at least one child aged 6–59 month, a total of 2,225 households would need to be approached for cohort recruitment. Given that 2500 households have been mapped in the study area, all households will be approached at baseline and again at 12 months during the LLIN census for recruitment of children into the cohort study. For the second cohort beginning in 2013, the parameters to be used are as above. The length of time the cohort is to be followed is decreased from 12 to 9 months, but this only increases the sample size by 6% to 1043. In the first cohort, we estimated a 5% refusal rate, but actually had a 12.4% refusal rate. An additional 5.3% of children were ineligible due to use of cotrimoxasole in HIV-exposed children, 3.4% had moved out of the study area by the time of enrollment, and 6.8% could not be located. Once enrolled...
