Sample Size Estimation Research Articles

Rising risks of hydroclimatic swings: A large ensemble study of dry and wet spell transitions in North America

Hydroclimatic swings between dry and wet spells are increasing globally, raising concerns due to their severe impacts on society, agriculture, infrastructure systems, and water resource management. These extremes and their projected changes are traditionally assessed in isolation, which can underestimate associated risks and future resilience and adaptation plans. This study investigates climate change projections of such lagged compound dry and wet spells across North America using a single model initial-condition large ensemble (SMILE). Three dry-wet spell indices are merged into an integrated indicator to provide a comprehensive perspective on changing risks. We apply an ensemble pooling approach using the Canadian Regional Climate Model version 4 Large Ensemble (CanRCM4-LE) to enhance sample size for index estimation. Results suggest that hydroclimatic swings across North America are expected to become more frequent and intensified in a warmer climate. Trends of spatial fraction and aggregation during transitions between dry and wet spells indicate future water resource management challenges. Hot spots for intensified transitions with higher frequency, such as Northern Central America, and the southern part of Eastern and Western North America, including Mexico and the state of California, overlap with larger spatial extent and higher aggregation during transitions. The seasonal analysis of spatial characteristics indicates winter may become wetter and summer drier at higher warming levels, potentially intensifying transitions between dry and wet extremes. Clustered hydroclimatic swings may lead to severe environmental, hydrological, and socio-economic consequences, necessitating appropriate mitigation and adaptation measures.

A Randomized Controlled Trial of Tenecteplase Versus Standard of Care for Minor Ischemic Stroke with Proven Occlusion (TEMPO-2): Rational and design of a multicenter, randomized open-label clinical trial.

Almost half of acute ischemic stroke patients present with mild symptoms and there are large practice variations in their treatment globally. Individuals with an intracranial occlusion who present with minor stroke are at an increased risk of early neurological deterioration and poor outcomes. Individual patient data meta-analysis in the subgroup of patients with minor deficits showed benefit of alteplase in improving outcomes; however, this benefit has not been seen with intravenous alteplase in published randomized trials. TEMPO-2 (A Randomized Controlled Trial of Tenecteplase Versus Standard of Care for Minor Ischemic Stroke With Proven Occlusion) is a prospective, open label with blinded outcome assessment, randomized controlled trial, designed to test the superiority of intravenous tenecteplase (0.25 mg/kg) over nonthrombolytic standard of care, with an estimated sample size of 1274 patients. Adult patients presenting with acute ischemic stroke with the National Institutes of Health Stroke Scale (NIHSS) ⩽ 5 and visible arterial occlusion or perfusion deficit within 12 h of onset are randomized to receive either tenecteplase (0.25 mg/kg) or standard of care. The primary outcome is return to baseline neurological functioning, measured by the modified Rankin scale (mRS) at 90 days. Safety outcomes include death and symptomatic hemorrhage (intra or extra-cranial). Other secondary outcomes include mRS 0-1, mRS 0-2, ordinal shift analysis of the mRS, partial, and full recanalization on follow-up computed tomography angiogram. Results of this trial will aid in determining whether there is benefit of using tenecteplase (0.25 mg/kg) in treating patients presenting with minor stroke who are at high risk of developing poor outcomes due to presence of an intracranial occlusion. Data will be available upon reasonable request.

Commentary: Customer-Perceived Innovation: Considerations for Financial Performance and Methodological Approaches

As firms increasingly uncover their activities to key stakeholders through various media, the perception of these activities is becoming more important for firm performance. Traditionally, access to industry-wide databases provides important metrics on customer perceptions of performance, such as customer satisfaction and brand equity. In addition, numerous studies have highlighted firms' innovation-related actions (e.g., R&D spending and patent counts) as critical metrics linked to their financial performance. Perceived Firm Innovation (PFI) emerges as a relatively new and under-studied metric with the potential to impact a firm's financial success. Keiningham et al. are among the pioneers in this area. This commentary views their article as a vital initial step in understanding PFI’s impact. Considering that the service sector accounts for over 70 percent of the GDP in any developed country, service innovation is a broad phenomenon. Given the breadth of the area, we point to the challenge of capturing this phenomenon with a single metric like PFI. We also discuss crucial methodological considerations for future research, including estimation methods, sample size, and financial metrics.

Data from the Paper Entitled “Application of a Bayesian Approach for Exploring the Impact of Syllable Frequency in Handwritten Picture Naming”

The data presented here comes from the Perret and Solier (2022) study. 30 participants handwrote labels for 150 black-and-white drawings. The experiment was carried out using the DmDx program. Response times and production errors were the two behavioral reported measures. DmDx scripts and data are available on the OSF platform (DOI: https://doi.org/10.17605/OSF.IO/GAZF3). These data should be useful for pre-testing to explore new hypotheses, as well as for methodological elements (e.g., sample size estimation, estimation of a priori distributions for Bayesian analyses).

Study protocol: Early neurological deterioration in patients with minor stroke, frequency, predictors, and outcomes in Vietnam single-centre study.

Early neurological deterioration (END) is progressive neurological deterioration with an increase in NIHSS score of 2 points or more in the first 72 hours from the onset of acute ischemic stroke. END increases the risk of poor clinical outcomes at day 90 of ischemic stroke. We will study the frequency, predictors, and outcomes of patients with END in a case-control study at a comprehensive stroke centre in Vietnam. of the design is a descriptive observational study, longitudinal follow-up of patients with minor stroke hospitalized at the Stroke Center of Bach Mai Hospital from December 1, 2023, to December 1, 2024. Minor stroke patients characterized by NIHSS score ≤ 5 hospitalized within 24 hours of symptom onset will be recruited. The estimated END rate is about 30%, relative accuracy ε = 0.11, 95% reliability, expected 5% of patients lost data or follow-up, and an estimated sample size of 779 patients. This study will help determine the END rate in patients with minor stroke and related factors, thereby building a prognostic model for END. Our study determined the END rate in patients with minor stroke in Vietnam and also proposed risk factors for minor stroke management and treatment.

ACETABULAR FRACTURES IN OLDER PATIENTS INTERVENTION TRIAL (ACEFIT): A FEASIBILITY TRIPLE-ARM RANDOMIZED CONTROLLED STUDY

Current treatment options for displaced acetabular fractures in elderly patients include non-surgical management, surgical fixation and surgical fixation with simultaneous hip replacement, the so-called “fix-and-replace”™. There remains a paucity of evidence to guide surgeons in decision making for these difficult injuries. The aim of this study was to assess the feasibility of performing an appropriately powered RCT between treatment options for acetabular fractures in older patients.This was an NIHR funded feasibility triple-arm RCT with participation from 7 NHS MTCs. Patients older than 60 were recruited if they had an acetabular fracture deemed sufficiently displaced for the treating surgeon to consider surgical fixation. Randomisation was performed on a 1:1:1 basis. The three treatment arms were non-surgical management, surgical fixation and fix-and-replace. Feasibility was assessed by willingness of patients to participate and clinicians to recruit, drop out rate, estimates of standard deviation to inform the sample size calculation for the full trial and completion rates to inform design of a future definitive trial. EQ-5D was the primary outcome measure at 6 months, OHS and Disability Rating Index were secondary outcome measures.Of 117 eligible patients, 60 were randomised whilst 50 declined study participation. Nine patients did not receive their allocated intervention. Analysis was performed on an intention to treat basis. During the study period 4 patients withdrew before final review, 4 patients died and 1 was lost to follow-up. The estimated sample size for a full scale study was calculated to be 1474 participants for an EQ-5D MCID of 0.06 with a power of 0.8.This feasibility study suggests a full scale trial would require international collaboration. This study also has provided observed safety data regarding mortality and morbidity for the fix-and-replace procedure to aid surgeons in the decision-making process when considering treatment options.

Efficacy of acupuncture-related therapy in the treatment of primary dysmenorrhea: A network meta-analysis of randomized controlled trials

ObjectivesIn order to compare and rank the most effective acupuncture therapy for primary dysmenorrhea and provide evidence-based medical support for clinical treatment of this disease. MethodsA comprehensive search was conducted on China National Knowledge Infrastructure (CNKI), Wanfang Database, Information Chinese Journal Service Platform (VIP), China Biomedical Literature Service System (SinoMed), PubMed, Web of Science, Embase, and Cochrane Library databases from their inception to May 1, 2023. The Cochrane Collaboration Risk of Bias Tool was used to evaluate bias risk, and the GeMTC package of Stata 15.1 software and R 4.3.1 software was used to perform network Meta-analysis. Results70 studies were included, including 5772 patients with primary dysmenorrhea, involving 25 kinds of acupuncture techniques commonly used in clinic. The quality of the included literature was low, most of them did not mention the registration information of clinical trial centers, and the specific sample size estimation method was unclear. Some literature did not explain the specific random method, distribution concealment and blindness, so there was a certain publication bias and small sample effect. Results showed that for improving the clinical effective rate, the top three treatments were salt-separated moxibustion, massotherapy + acupoint patching, acupuncture + heat-sensitive moxibustion. In terms of reducing the visual analogue scale(VAS), the top three treatments were massotherapy + acupoint patching, acupuncture + acupoint patching and warm acupuncture. In terms of alleviating cox menstrual symptom scale (CMSS), the top three treatments were acupuncture + acupoint patching, acupoint patching and point embedding. In relieving TCM symptom score, the top three treatments were acupoint patching + heat-sensitive moxibustion, acupoint patching and moxibustion. ConclusionDifferent acupuncture therapies have more advantages than oral analgesics in improving the clinical effective rate, reducing VAS score, reducing CMSS score, and alleviating TCM symptom score. Among them, massage therapy + acupoint patching, acupuncture + acupoint patching and acupoint patching may be the best solutions for the treatment of primary dysmenorrhea. However, more large-sample, multi-center and high-quality randomized controlled trials are needed to demonstrate.

Perfusion outcomes with near-infrared indocyanine green imaging system in laparoscopic total mesorectal excision for mid-rectal or low-rectal cancer (POSTER): a study protocol

IntroductionAnastomotic leakage (AL) is defined as the failure of complete healing or disruption of the anastomosis subsequent to rectal cancer surgery, resulting in the extravasation of intestinal contents into the...

Diagnostic Accuracy of Abdominal Point of Care Ultrasound in Primary Care: Study Design and Protocol.

The aim of this study is to estimate the diagnostic accuracy of abdominal point of care ultrasound (POCUS) performed by family physicians (FPs) in primary care (PC), in comparison with the findings in the medical record (MR) at 12 months of follow-up. This study is conducted entirely in PC healthcare centers in Spain. Abdominal ultrasound scans performed by FPs (selected on the basis of their ultrasound knowledge and experience) are compared with the findings, or not, in the patient's MR after a 12-month follow-up period. The study will involve 100 FPs in Spain and an estimated sample size of 1334 patients who are to undergo abdominal POCUS at the indication of their physician. The results of the abdominal POCUS will be collected and compared with the findings of the MR. This comparison will be performed by another physician of the research team, different from their FP after one year of follow-up.The diagnostic accuracy of abdominal POCUS has been addressed in the hospital setting but not in PC. This lack of evidence can begin to be resolved with studies such as the one we present, designed for unselected populations such as those treated in PC and taking the patient's MR as the gold standard, which will allow us to make comparisons with the patient's clinical course.

Sample size estimation for AQP4-IgG seropositive optic neuritis: Retinal damage detection by optical coherence tomography.

We recruited four aquaporin-4 seropositive optic neuritis patients (five eyes) who received glucocorticoid treatment and underwent optical coherence tomography examination. Baseline medians of the macular ganglion cell layer plus inner plexiform layer (mGCIPL) thickness and volume for the eye of interest were 79.67 µm (73.664 ± 18.497 µm) and 0.58 mm3 (0.534 ± 0.134 mm3), respectively. At 2 months, the medians of the mGCIPL thickness and volume were 60.00 µm (51.576 ± 12.611 µm) and 0.44 mm3 (0.376 ± 0.091 mm3), respectively. At 6 months, the medians of the mGCIPL thickness and volume were 59.55 µm (46.288 ± 11.876 µm) and 0.44 mm3 (0.336 ± 0.084 mm3), respectively. Sample size estimate was achieved using two methods based on the mGCIPL thickness and volume data, with five effect sizes considered. The estimate based on the mGCIPL volume showed that 206 patients were needed at the 6-month follow-up; the power was 80% and effect size was 20%. In conclusion, this study detected retinal damage in aquaporin-4 seropositive optic neuritis patients by optical coherence tomography, and estimated the sample size for two-sample parallel designed clinical trials using two methods.

Rationale and study design for an Individualized PeriopeRative Open lung VEntilatory approach in Emergency Abdominal Laparotomy/scopy: study protocol for a prospective international randomized controlled trial

BackgroundPostoperative pulmonary complications (PPC) are the most frequent postoperative complications, with an estimated prevalence in elective surgery ranging from 20% in observational cohort studies to 40% in randomized clinical trials. However, the prevalence of PPCs in patients undergoing emergency abdominal surgery is not well defined. Lung-protective ventilation aims to minimize ventilator-induced lung injury and reduce PPCs. The open lung approach (OLA), which combines recruitment manoeuvres (RM) and positive end-expiratory pressure (PEEP) titration, aims to minimize areas of atelectasis and the development of PPCs; however, there is no conclusive evidence in the literature that OLA can prevent PPCs. The purpose of this study is to compare an individualized perioperative OLA with conventional standardized lung-protective ventilation in patients undergoing emergency abdominal surgery with clinical signs of intraoperative lung collapse. MethodsRandomized international clinical trial to compare an individualized perioperative OLA (RM plus individualized PEEP and individualized postoperative respiratory support) with conventional lung-protective ventilation (standard PEEP of 5 cmH2O and conventional postoperative oxygen therapy) in patients undergoing emergency abdominal surgery with clinical signs of lung collapse. Patients will be randomised to open-label parallel groups. The primary outcome is any severe PPC during the first 7 postoperative days, including: acute respiratory failure, pneumothorax, weaning failure, acute respiratory distress syndrome, and pulmonary infection. The estimated sample size is 732 patients (366 per group). The final sample size will be readjusted during the interim analysis. DiscussionThe Individualized Perioperative Open-lung Ventilatory Strategy in emergency abdominal laparotomy (iPROVE-EAL) is the first multicentre, randomized, controlled trial to investigate whether an individualized perioperative approach prevents PPCs in patients undergoing emergency surgery.

The 3 + 3 design in dose-finding studies with small sample sizes: Pitfalls and possible remedies.

In the last few years, numerous novel designs have been proposed to improve the efficiency and accuracy of phase I trials to identify the maximum-tolerated dose (MTD) or the optimal biological dose (OBD) for noncytotoxic agents. However, the conventional 3+3 approach, known for its and poor performance, continues to be an attractive choice for many trials despite these alternative suggestions. The article seeks to underscore the importance of moving beyond the 3+3 design by highlighting a different key element in trial design: the estimation of sample size and its crucial role in predicting toxicity and determining the MTD. We use simulation studies to compare the performance of the most used phase I approaches: 3+3, Continual Reassessment Method (CRM), Keyboard and Bayesian Optimal Interval (BOIN) designs regarding three key operating characteristics: the percentage of correct selection of the true MTD, the average number of patients allocated per dose level, and the average total sample size. The simulation results consistently show that the 3+3 algorithm underperforms in comparison to model-based and model-assisted designs across all scenarios and metrics. The 3+3 method yields significantly lower (up to three times) probabilities in identifying the correct MTD, often selecting doses one or even two levels below the actual MTD. The 3+3 design allocates significantly fewer patients at the true MTD, assigns higher numbers to lower dose levels, and rarely explores doses above the target dose-limiting toxicity (DLT) rate. The overall performance of the 3+3 method is suboptimal, with a high level of unexplained uncertainty and significant implications for accurately determining the MTD. While the primary focus of the article is to demonstrate the limitations of the 3+3 algorithm, the question remains about the preferred alternative approach. The intention is not to definitively recommend one model-based or model-assisted method over others, as their performance can vary based on parameters and model specifications. However, the presented results indicate that the CRM, Keyboard, and BOIN designs consistently outperform the 3+3 and offer improved efficiency and precision in determining the MTD, which is crucial in early-phase clinical trials.

Survey on the awareness rate of mpox knowledge and related factors among men who have sex with men in China

Objective: To understand the awareness rate of mpox knowledge and related factors among men who have sex with men (MSM) in China. Methods: The survey was conducted among men aged ≥18 years who had sex with men in the past year, using the convenience sampling method. The estimated sample size was 4 312. With the assistance of social organizations of MSM in 30 provinces in China, an online questionnaire survey was conducted using anonymous self-designed questionnaires powered by www.wjx.cn during 10-14 August 2023 to collect information on socio-demographic characteristics, awareness of mpox knowledge, travel history, and sexual behaviors of the respondents. The software SAS 9.4 was used for statistical analysis. Results: There were 7 725 respondents, and the age of the respondents was (31.6±9.0) years. The results revealed that the awareness rate of mpox knowledge was 50.1% (3 872/7 725). The main routes to acquire mpox knowledge were mainly new media, including WeChat, Weibo, TikTok, and Blued social software (88.4%,6 827/7 725), while official media report was the most trusted way to acquire related knowledge (79.3%,6 129/7 725). Multivariate logistic regression analysis revealed that the factors indicated a higher awareness rate of mpox knowledge, including living in the western region or the eastern region, people were over 26 years old in early adulthood and midlife, college-educated or with higher degrees, living in towns or urban periphery, being in homosexual or in bisexual relationships, 1-5 homosexual times/months in the past 3 months, knowing their HIV infection status, paying attention to mpox knowledge very often, occasionally or rarely, and convenient ways to acquire mpox knowledge from new media and social organizations. Conclusions: The awareness rate of mpox knowledge was low among MSM in China. Efforts should be made to improve the awareness rate of mpox knowledge among those who are young, less educated, or sexually active, with targeted health education via new media and social organizations.

Bayesian Multivariate Logistic Regression for Superiority and Inferiority Decision-Making under Observable Treatment Heterogeneity

The effects of treatments may differ between persons with different characteristics. Addressing such treatment heterogeneity is crucial to investigate whether patients with specific characteristics are likely to benefit from a new treatment. The current paper presents a novel Bayesian method for superiority decision-making in the context of randomized controlled trials with multivariate binary responses and heterogeneous treatment effects. The framework is based on three elements: a) Bayesian multivariate logistic regression analysis with a Pólya-Gamma expansion; b) a transformation procedure to transfer obtained regression coefficients to a more intuitive multivariate probability scale (i.e., success probabilities and the differences between them); and c) a compatible decision procedure for treatment comparison with prespecified decision error rates. Procedures for a priori sample size estimation under a non-informative prior distribution are included. A numerical evaluation demonstrated that decisions based on a priori sample size estimation resulted in anticipated error rates among the trial population as well as subpopulations. Further, average and conditional treatment effect parameters could be estimated unbiasedly when the sample was large enough. Illustration with the International Stroke Trial dataset revealed a trend toward heterogeneous effects among stroke patients: Something that would have remained undetected when analyses were limited to average treatment effects.

Navigating Sample Size Estimation for Qualitative Research

Abstract There are well-established rules and methods about sample size estimation in quantitative research approaches. However, qualitative research approaches justify very little about sample size estimation principles and largely depend on subjective judgements and arbitrariness. Contrarily, an adequate sample size is essential for a study to address the core elements of validity and credibility in qualitative research too such as rigor, trustworthiness, conformability and acceptance. Therefore, this review was carried out to explain the available methods to estimate sample size for qualitative studies. After conducting a thorough literature review, we discovered related articles that explore the estimation of sample size for qualitative studies. By examining these findings and integrating the information with our personal experience for estimation of sample size in the field of qualitative studies, we have produced an all-encompassing narrative review. After an in-depth literature search, four different approaches were described in this paper to answer the question of how to estimate sample size in qualitative studies. The four approaches described in this paper are (a) rules of thumb, (b) conceptual models, (c) concept of saturation and (d) statistics-based methods for sample size estimation in qualitative research. The paper presents four methods for estimating sample size in qualitative studies and simplifies the statistical approach for saturation calculation in qualitative studies. Yet, it is vital to responsibly integrate these methods, acknowledging their limitations and maintaining the importance of sample size estimation in qualitative studies.

Sample size planning for rank-based multiple contrast tests.

Rank methods are well-established tools for comparing two or multiple (independent) groups. Statistical planning methods for the computing the required sample size(s) to detect a specific alternative with predefined power are lacking. In the present paper, we develop numerical algorithms for sample size planning of pseudo-rank-based multiple contrast tests. We discuss the treatment effects and different ways to approximate variance parameters within the estimation scheme. We further compare pairwise with global rank methods in detail. Extensive simulation studies show that the sample size estimators are accurate. A real data example illustrates the application of themethods.

IJCM_239A: Health seeking behaviour and its determinants among adult population: a comparative study between urban and rural area of western Uttar Pradesh

Background: Health seeking behaviour is remedial action undertaken by individuals for their health problem or illness. It may vary in rural and urban areas and with different sociodemographic variables. To provide adequate health care services to the community, it is essential to know about their health seeking behaviour and its determinants. Objectives: 1. To describe health seeking behaviour and determine various factors influencing it among adult population in rural and urban areas of Meerut. 2. To assess their awareness on health insurance policies. Methodology: A community based cross-sectional study among adult population was conducted in rural and urban areas of Meerut, Uttar Pradesh from 1st July to 30th September, 2023. Estimated sample size was 200 each from rural and urban areas. Study participants were selected by multistage sampling method. Data collection was done by a pre-tested and semi-structured questionnaire. Results: Among 400 study participants, 91% in urban and 60% in rural areas showed positive health seeking behaviour. In both areas, most of them preferred public health facilities. Major factor affecting health seeking behaviour in rural area was convenience of approach, while that in urban area was low cost of treatment. Majority rural population was unaware of health insurance policies, while almost half of urban population were aware of it. Conclusion: Positive health seeking behaviour and awareness on health insurance policies are seen more in urban area than rural. Awareness programmes on positive health seeking behaviour and health insurance are essential, especially in rural areas.

IJCM_139A: Prevalence of Hypertension among the Urban Adolescents: A Prospective Robust Survey

Background: Childhood Blood pressure (BP) is a strong indicator of adult blood pressure (BP), hence, early detection and intervention is important in adolescents. Studies from China and United States have shown BP in adolescents to be variable and has indeed outlined the need to obtain three separate readings before diagnosing hypertension. There is a dearth of this documentation of blood pressure readings at different occasions in adolescents from the Indian settings. Objective: To determine the prevalence of elevated blood pressure and hypertension among 10-18 years school going adolescents. Methodology: This was a prospective longitudinal study conducted across the five urban schools of Bhopal amongst 10-18 years adolescents. Multistage cluster sampling was adopted to cover an estimated sample size of 780 over a one-year period. Those adolescents who were present during the time of survey following their parent’s consent were included. Those who were absent at two occasions during the follow up visit and those who were diagnosed with secondary hypertension were excluded. Those who satisfied the study criteria were given a self-administered WHO STEPS questionnaire initially. The height and weight were obtained followed by measurement of three auscultatory BP readings at an interval of one week apart (American Academy of paediatrics (Updated guidelines 2017) in each student. The R shiny app calculator was used for classifying hypertension in adolescents. The study was funded by AIIMS Bhopal and ethical clearance was obtained from the institute. Results: The overall prevalence of elevated BP/hypertension in adolescents was 1.4% (N=795). There was a positive correlation between BMI and (SBP & DBP). The generalised linear mixed model showed age, BMI and boys to have higher predisposition to elevated BP/HTN. There was a significant difference in the SBP and DBP between the 3 visits stratified by sex. Also, the change in SBP and DBP between the readings were significant by ANOVA. Conclusion: The prevalence of elevated BP/HTN was 1.4% with girls being protective for girls. The study cannot be generalised as it is limited to the Government schools with relatively small sample size.

Sample size estimation for stratified cluster randomization trial with survival endpoint.

Cluster randomization trials with survival endpoint are predominantly used in drug development and clinical care research when drug treatments or interventions are delivered at a group level. Unlike conventional cluster randomization design, stratified cluster randomization design is generally considered more effective in reducing the impacts of imbalanced baseline prognostic factors and varying cluster sizes between groups when these stratification factors are adopted in the design. Failure to account for stratification and cluster size variability may lead to underpowered analysis and inaccurate sample size estimation. Apart from the sample size estimation in unstratified cluster randomization trials, there are no development of an explicit sample size formula for survival endpoint when a stratified cluster randomization design is employed. In this article, we present a closed-form sample size formula based on the stratified cluster log-rank statistics for stratified cluster randomization trials with survival endpoint. It provides an integrated solution for sample size estimation that account for cluster size variation, baseline hazard heterogeneity, and the estimated intracluster correlation coefficient based on the preliminary data. Simulation studies show that the proposed formula provides the appropriate sample size for achieving the desired statistical power under various parameter configurations. A real example of a stratified cluster randomization trial in the population with stable coronary heart disease is presented to illustrate our method.

An Assessment of Knowledge of Health Workers in Hospitals of Southern Province of Zambia, Towards Marketing of Breastmilk Substitutes Regulations

Background: Optimal breastfeeding practices have been undermined by unregulated marketing of breastmilk substitutes worldwide, resulting in about 820, 000 preventable deaths per year among children under five years globally. The International Code (thereafter called the Code) of Marketing Breastmilk Substitutes was developed by the World Health Organization, the United Nations Children's Emergency Fund and stakeholders to protect, promote and support breastfeeding. Member states are expected to domesticate the Code. Zambia enacted Statutory Instrument No. 48 of 2006, Regulations on marketing of breastmilk substitutes. Successful implementation calls for a good level of knowledge among health workers. However, we found no study, in Zambia on the marketing of breastmilk substitutes Regulations. Objective: To assess the level of knowledge of health workers on Zambia’s Regulations on marketing breastmilk substitutes. Study design and methods: A mixed-method cross-sectional study. Data was collected online and in person. Probability proportional to size calculation yielded an estimated sample size of 384. A sample size of 410 after adjusting upwards assumed non-response rate, 401 participants responded. Quantitative data were analyzed using the Social Statistical Package for Social Sciences. Knowledge results were described by frequencies and percentages. Tests were conducted using univariable linear and multivariable regression analysis. A p-value of <0.05 was considered statistically significant. Thematic analysis was used for qualitative data analysis. Results: The mean overall level of good knowledge of health workers was 0.339 or 34% [SE: 0.0221; 95%CI: (0.2959, 0.3828)]. Conclusion: The level of knowledge among health workers in hospitals of Southern Province, Zambia was poor.