OBJECTIVE: To compare different methods for calculating sample size based on confidence interval estimation for a single proportion with different event incidences and precisions. METHODS: We compared 7 methods, namely Wald, AgrestiCoull add z2, Agresti-Coull add 4, Wilson Score, Clopper-Pearson, Mid-p, and Jefferys, for confidence interval estimation for a single proportion. The sample size was calculated using the search method with different parameter settings (proportion of specified events and half width of the confidence interval [ω=0.05, 0.1]). With Monte Carlo simulation, the estimated sample size was used to simulate and compare the width of the confidence interval, the coverage of the confidence interval and the ratio of the noncoverage probability. RESULTS: For a high accuracy requirement (ω =0.05), the Mid-p method and Clopper Pearson method performed better when the incidence of events was low (P < 0.15). In other settings, the performance of the 7 methods did not differ significantly except for a poor symmetry of the Wald method. In the setting of ω=0.1 with a very low p (0.01-0.05), failure of iteration occurred with nearly all the methods except for the Clopper-Pearson method. CONCLUSION Different sample size determination methods based on confidence interval estimation should be selected for single proportions with different parameter settings.
Confidence Intervals ; Sample Size ; Computer Simulation ; Monte Carlo Method ; Probability

Objective: To systematically introduce the design of case-cohort study and the statistical methods of relative risk estimation and their application in the design. Methods: First, we introduced the basic principles of case-cohort study design. Secondly, Prentice's method, Self-Prentice method and Barlow method were described in the weighted Cox proportional hazard regression models in detail, finally, the data from the Shanghai Women's Health Study were used as an example to analyze the association between obesity and liver cancer incidence in the full cohort and case-cohort sample, and the results of parameters from each method were compared. Results: Significant association was observed between obesity and risk for liver cancer incidence in women in both the full cohort and the case-cohort sample. In the Cox proportional hazard regression model, the partial regression coefficients of the full cohort and the case-cohort sample fluctuated with the adjustment of confounding factors, but the hazard ratio estimates of them were close. There was a difference in the standard error of the partial regression coefficient between the full cohort and the case-cohort sample. The standard error of the partial regression coefficient of the case-cohort sample was larger than that of the full cohort, resulting in a wider 95% confidence interval of the relative risk. In the weighted Cox proportional hazard regression model, the standard error of the partial regression coefficient of Prentice's method was closer to the parameter estimates from full cohort than Self-Prentice method and Barlow method, and the 95% confidence interval of hazard ratio was closer to that of the full cohort. Conclusions: Case-cohort design could yield parameter results closer to the full cohort by collecting and analyzing data from sub-cohort members and patients with the disease, and reduce sample size and improve research efficiency. The results suggested that Prentice's method would be preferred in case-cohort design.
China/epidemiology* ; Cohort Studies ; Female ; Humans ; Proportional Hazards Models ; Risk ; Sample Size

Sample size computations, which should be done at the planning stage of the study, are necessary for research to estimate a population parameter or test a hypothesis. For causal analysis of observational databases, sample size computations are generally not needed. Post-hoc power analyses, which are typically done with non-significant findings, should not be performed since reporting post-hoc power is nothing more than reporting p values differently. While sample size calculations are typically based on the tradition of significance testing, sample size calculations based on precision are feasible – if not preferred – alternatives. Sample size calculations depend on several factors such as the study objective, scale of measurement of the outcome variable, study design, and sampling design. Computing for sample size is not as straightforward as presented in textbooks but specific strategies may be resorted to in the face of challenges and constraints.
Sample Size ; Power, Psychological

The appropriate sample size estimation is very important in the design of clinical trials. However, insufficient or inappropriate sample size estimation is still a prominent problem in the currently published acupuncture and moxibustion clinical trials. At present, the superiority test, non-inferiority test and equivalence test have been widely used in acupuncture and moxibustion clinical trials. This article focuses on the application, calculation methods and PASS11 software using of these three hypothesis test types. In view of the problems in the estimation of sample size in acupuncture and moxibustion clinical trials, the particularity of sample size estimation in acupuncture and moxibustion is summarized from the aspects of parameter setting, ratio of intervention group and control group, and multi-group comparison, in order to guide acupuncture clinical researchers to correctly estimate sample size when conducting clinical trials.
Acupuncture ; Acupuncture Therapy ; Clinical Trials as Topic ; Moxibustion ; Sample Size

BACKGROUND: Pulmonary hypertension (PH) is common in patients with idiopathic pulmonary fibrosis (IPF) and is associated with poor outcomes. This study was performed to determine the clinical efficacy of PH-specific therapeutic agents for IPF patients.METHODS: We performed a systematic review and meta-analysis using MEDLINE, EMBASE, and the Cochrane Central Register. We searched randomized controlled trials (RCTs) without language restriction until November 2018. The primary outcome was all-cause mortality to end of study.RESULTS: We analyzed 10 RCTs involving 2,124 patients, 1,274 of whom received PH-specific agents. In pooled estimates, the use of PH-specific agents was not significantly associated with reduced all-cause mortality to end of study compared with controls (hazard ratio, 0.99; 95% confidence interval [CI], 0.92, 1.06; P = 0.71; I² = 30%). When we performed subgroup analyses according to the type of PH-specific agent, sample size, age, forced vital capacity, diffusion lung capacity, and the extent of honeycombing, PH-specific agents also showed no significant association with a reduction in all-cause mortality. A small but significant improvement in quality of life, measured using the St. George Respiratory Questionnaire total score, was found in the PH-specific agent group (mean difference, −3.16 points; 95% CI, −5.34, −0.97; P = 0.005; I² = 0%). We found no significant changes from baseline in lung function, dyspnea, or exercise capacity. Serious adverse events were similar between the two groups.CONCLUSION: Although PH-specific agents provided small health-related quality-of-life benefits, our meta-analysis provides insufficient evidence to support their use in IPF patients.
Diffusion ; Dyspnea ; Humans ; Hypertension, Pulmonary ; Idiopathic Pulmonary Fibrosis ; Lung ; Lung Volume Measurements ; Mortality ; Quality of Life ; Sample Size ; Treatment Outcome ; Vasodilator Agents ; Vital Capacity

PURPOSE: To compare the accuracy of three intraocular lens (IOL) power calculation formulas (SRK/T, Barrett Universal II, and T2) in cataract surgery patients.METHODS: In total, 73 eyes of 73 patients who underwent uneventful cataract surgery were retrospectively reviewed. IOL power was determined using SRK/T, Barrett Universal II, and T2 preoperatively. The findings were compared with the actual refractive outcome to obtain the prediction error. The mean prediction error (ME) and mean absolute error (MAE) of each formula were compared. The MAE was defined as the difference between the postoperative spherical equivalence (SE) and the preoperatively predicted SE. The ME and MAE of each formula 3 months after surgery were compared with preoperatively predicted SE. Eyes were classified into subgroups based on axial length (AL) and average keratometry (K).RESULTS: The ME and MAE for the three formulas were SRK/T [−0.08 ± 0.45 diopters (D) and 0.35 ± 0.40 D, respectively], Barrett Universal II (−0.01 ± 0.44 D and 0.33 ± 0.30 D, respectively), and T2 (0.04 ± 0.45 D and −0.34 ± 0.30 D, respectively), but no statistically significant differences were detected. Similar results were obtained in groups with a long AL or a large average K. In groups with an AL ≥ 26 mm or with an average K ≥ 47 D, the Barrett Universal II formula yielded the smallest standard deviation and a ME closest to zero, but these differences were not statistically significant.CONCLUSIONS: No significant differences were observed between the three formulas regarding ME or MAE. However, recent formulas such as the Barrett Universal II could provide certain benefits in predicting IOL power for patients with a long AL (> 26 mm) or larger average K. Further research with a larger sample size is recommended for more evaluation.
Cataract ; Humans ; Lenses, Intraocular ; Retrospective Studies ; Sample Size

sample size, it suggests that maintenance carboplatin-PLD chemotherapy could improve PFS in advanced ovarian cancer.]]>
Arm ; Asian Continental Ancestry Group ; Bias (Epidemiology) ; Carboplatin ; Disease-Free Survival ; Doxorubicin ; Drug Therapy ; Follow-Up Studies ; Humans ; Maintenance Chemotherapy ; Ovarian Neoplasms ; Prognosis ; Quality of Life ; Recombination, Genetic ; Sample Size

OBJECTIVE: To explore the relationship between sample size in the groups and statistical power of ANOVA and Kruskal-Wallis test with an imbalanced design. METHODS: The sample sizes of the two tests were estimated by SAS program with given parameter settings, and Monte Carlo simulation was used to examine the changes in power when the total sample size varied or remained fixed. RESULTS: In ANOVA, when the total sample size was fixed, increasing the sample size in the group with a larger mean square error improved the statistical power, but an excessively large difference in the sample sizes between groups led to reduced power. When the total sample size was not fixed, a larger mean square error in the group with increased sample size was associated with a greater increase of the statistical power. In Kruskal-wallis test, when the total sample size was fixed, increasing the sample size in groups with large mean square errors increased the statistical power irrespective of the sample size difference between the groups; when total sample size was not fixed, a larger mean square error in the group with increased sample size resulted in an increased statistical power, and the increment was similar to that for a fixed total sample size. CONCLUSIONS The relationship between statistical power and sample size in groups is affected by the mean square error, and increasing the sample size in a group with a large mean square error increases the statistical power. In Kruskal-Wallis test, increasing the sample size in a group with a large mean square error is more cost- effective than increasing the total sample size to improve the statistical power.
Computer Simulation ; Models, Statistical ; Monte Carlo Method ; Sample Size

OBJECTIVE: To explore the application and advantages of conditional inference forest in survival analysis. METHODS: We used simulated experiment and actual data to compare the predictive performance of 4 models, including Coxproportional hazards model, accelerated failure time model, random survival forest model and conditional inference forest model based on their Brier scores. RESULTS: Simulation experiment suggested that both of the two forest models had more accurate and robust predictive performance than the other two regression models. Conditional inference forest model was superior to the other models in analyzing time-to-event data with polytomous covariates, collinearity or interaction, especially for a large sample size and a high censoring rate. The results of actual data analysis demonstrated that conditional inference forest model had the best predictive performance among the 4 models. CONCLUSIONS Compared with the commonly used survival analysis methods, conditional inference forest model performs better especially when the data contain polytomous covariates with collinearity and interaction.
Data Analysis ; Proportional Hazards Models ; Sample Size ; Survival Analysis

BACKGROUND: Metabolic syndrome (MetS) is relatively common worldwide and an important risk factor for cardiovascular diseases. It is closely linked to arterial stiffness of the carotid artery. However, the association of MetS with the safety of carotid revascularization has been rarely studied. The aim of this study was to observe the current status of MetS and its components in Chinese carotid revascularized patients, and investigate the impact on major adverse clinical events (MACEs) after carotid endarterectomy (CEA) or carotid artery stenting (CAS). METHODS: From January 2013 to December 2017, patients undergoing CEA or CAS in the Neurosurgery Department of Xuanwu Hospital were retrospectively recruited. The changes in prevalence of MetS and each component with time were investigated. The primary outcome was 30-day post-operative MACEs. Univariable and multivariable analyses were performed to identify the impact of MetS on CEA or CAS. RESULTS: A total of 2068 patients who underwent CEA (766 cases) or CAS (1302 cases) were included. The rate of MetS was 17.9%; the prevalence rate of MetS increased with time. The occurrence rate of MACEs in CEA was 3.4% (26 cases) and in CAS, 3.1% (40 cases). There was no statistical difference between the two groups (3.4% vs. 3.1%, P = 0.600). For CEA patients, univariate analysis showed that the MACE (+) group had increased diabetes history (53.8% vs. 30.9%, P = 0.014) and MetS (34.6% vs. 15.8%, P = 0.023). For CAS patients, univariate analysis showed that the MACE (+) group had increased coronary artery disease history (40.0% vs. 21.6%, P = 0.006) and internal carotid artery tortuosity (67.5%% vs. 37.6%, P < 0.001). Furthermore, the MACE (+) group had higher systolic blood pressure (143.38 ± 22.74 vs. 135.42 ± 17.17 mmHg, P = 0.004). Multivariable analysis showed that the influencing factors for MACEs in CEA included history of diabetes (odds ratio [OR] = 2.345; 95% confidence interval [CI] = 1.057-5.205; P = 0.036) and MetS (OR = 2.476; 95% CI = 1.065-5.757; P = 0.035). The influencing factors for MACEs in CAS included systolic blood pressure (OR = 1.023; 95% CI = 1.005-1.040; P = 0.010), coronary artery disease (OR = 2.382; 95% CI = 1.237-4.587; P = 0.009) and internal carotid artery tortuosity (OR = 3.221; 95% CI = 1.637-6.337; P = 0.001). CONCLUSIONS The prevalence rate of MetS increased with time in carotid revascularized patients. MetS is a risk for short-term MACEs after CEA, but not CAS.
Carotid Arteries/surgery* ; Carotid Stenosis/surgery* ; China/epidemiology* ; Endarterectomy, Carotid/adverse effects* ; Humans ; Metabolic Syndrome/epidemiology* ; Retrospective Studies ; Risk Factors ; Sample Size ; Stents/adverse effects* ; Stroke ; Time Factors ; Treatment Outcome