NLM’s hierarchical vocabulary, known as MeSH (Medical Subject Heading), is utilized for the purpose of indexing articles in PubMed.

Evidence-based medicine involves four basic steps: developing a focused clinical question, searching for the best evidence, critically appraising the evidence, and applying the evidence and evaluating the outcome. When developing a question, it is important to understand the difference between background and foreground questions. Background questions are general questions about conditions, illnesses, syndromes, and pathophysiology, while foreground questions are more often about issues of care. The PICO system is often used to define the components of a foreground question: patient group of interest, intervention of interest, comparison, and primary outcome.

When searching for evidence, it is important to have a basic understanding of the types of evidence and sources of information. Scientific literature is divided into two basic categories: primary (empirical research) and secondary (interpretation and analysis of primary sources). Unfiltered sources are large databases of articles that have not been pre-screened for quality, while filtered resources summarize and appraise evidence from several studies.

There are several databases and search engines that can be used to search for evidence, including Medline and PubMed, Embase, the Cochrane Library, PsycINFO, CINAHL, and OpenGrey. Boolean logic can be used to combine search terms in PubMed, and phrase searching and truncation can also be used. Medical Subject Headings (MeSH) are used by indexers to describe articles for MEDLINE records, and the MeSH Database is like a thesaurus that enables exploration of this vocabulary.

Understanding Hazard Ratio in Survival Analysis

Survival analysis is a statistical method used to analyze the time it takes for an event of interest to occur, such as death of disease progression. In this type of analysis, the hazard ratio (HR) is a commonly used measure that is similar to the relative risk but takes into account the fact that the risk of an event may change over time.

The hazard ratio is particularly useful in situations where the risk of an event is not constant over time, such as in medical research where patients may have different survival times of disease progression rates. It is a measure of the relative risk of an event occurring in one group compared to another, taking into account the time it takes for the event to occur.

For example, in a study comparing the survival rates of two groups of cancer patients, the hazard ratio would be used to compare the risk of death in one group compared to the other, taking into account the time it takes for the patients to die. A hazard ratio of 1 indicates that there is no difference in the risk of death between the two groups, while a hazard ratio greater than 1 indicates that one group has a higher risk of death than the other.

Overall, the hazard ratio is a useful tool in survival analysis that allows researchers to compare the risk of an event occurring between different groups, taking into account the time it takes for the event to occur.

PubMed receives new references faster than MEDLINE because they do not need to undergo indexing, such as adding MeSH headings and checking tags. While an increasing number of MEDLINE citations have a link to the complete article, not all of them do. Since 2010, Embased has included all MEDLINE citations in its database, but it does not have all citations from before that year.

Evidence-based medicine involves four basic steps: developing a focused clinical question, searching for the best evidence, critically appraising the evidence, and applying the evidence and evaluating the outcome. When developing a question, it is important to understand the difference between background and foreground questions. Background questions are general questions about conditions, illnesses, syndromes, and pathophysiology, while foreground questions are more often about issues of care. The PICO system is often used to define the components of a foreground question: patient group of interest, intervention of interest, comparison, and primary outcome.

When searching for evidence, it is important to have a basic understanding of the types of evidence and sources of information. Scientific literature is divided into two basic categories: primary (empirical research) and secondary (interpretation and analysis of primary sources). Unfiltered sources are large databases of articles that have not been pre-screened for quality, while filtered resources summarize and appraise evidence from several studies.

There are several databases and search engines that can be used to search for evidence, including Medline and PubMed, Embase, the Cochrane Library, PsycINFO, CINAHL, and OpenGrey. Boolean logic can be used to combine search terms in PubMed, and phrase searching and truncation can also be used. Medical Subject Headings (MeSH) are used by indexers to describe articles for MEDLINE records, and the MeSH Database is like a thesaurus that enables exploration of this vocabulary.

The Kaplan-Meier plot is the most effective graphical representation of survival probability. It presents the overall likelihood of an individual’s survival over time from a baseline, and the comparison of two lines on the plot can indicate whether there is a survival advantage. To determine if the distinction between the two groups is significant, a log rank test can be employed.

An NNT of 1 indicates that every patient who receives the treatment experiences a positive outcome, while no patient in the control group experiences the same outcome. This represents an ideal outcome.

Measures of Effect in Clinical Studies

When conducting clinical studies, we often want to know the effect of treatments of exposures on health outcomes. Measures of effect are used in randomized controlled trials (RCTs) and include the odds ratio (of), risk ratio (RR), risk difference (RD), and number needed to treat (NNT). Dichotomous (binary) outcome data are common in clinical trials, where the outcome for each participant is one of two possibilities, such as dead of alive, of clinical improvement of no improvement.

To understand the difference between of and RR, it’s important to know the difference between risks and odds. Risk is a proportion that describes the probability of a health outcome occurring, while odds is a ratio that compares the probability of an event occurring to the probability of it not occurring. Absolute risk is the basic risk, while risk difference is the difference between the absolute risk of an event in the intervention group and the absolute risk in the control group. Relative risk is the ratio of risk in the intervention group to the risk in the control group.

The number needed to treat (NNT) is the number of patients who need to be treated for one to benefit. Odds are calculated by dividing the number of times an event happens by the number of times it does not happen. The odds ratio is the odds of an outcome given a particular exposure versus the odds of an outcome in the absence of the exposure. It is commonly used in case-control studies and can also be used in cross-sectional and cohort study designs. An odds ratio of 1 indicates no difference in risk between the two groups, while an odds ratio >1 indicates an increased risk and an odds ratio <1 indicates a reduced risk.

The prevalence of the disease is 20% as there are currently 200 cases out of a total population of 1000.

When investigating prognosis, the hierarchy of study designs starts with a systematic review of cohort studies, followed by a cohort study, follow-up of untreated patients from randomized controlled trials, case series, and expert opinion. The strength of evidence provided by a study depends on its ability to minimize bias and maximize attribution. The Agency for Healthcare Policy and Research hierarchy of study types is widely accepted as reliable, with systematic reviews and meta-analyses of randomized controlled trials at the top, followed by randomized controlled trials, non-randomized intervention studies, observational studies, and non-experimental studies.

Cost minimisation analysis is employed to compare net costs when the observed effects of health care interventions are similar. To conduct this analysis, it is necessary to have clinical evidence that demonstrates the differences in health effects between alternatives are negligible of insignificant. This approach is commonly used by institutions like the National Institute for Health and Care Excellence (NICE).

Understanding Hypothesis Testing in Statistics

In statistics, it is not feasible to investigate hypotheses on entire populations. Therefore, researchers take samples and use them to make estimates about the population they are drawn from. However, this leads to uncertainty as there is no guarantee that the sample taken will be truly representative of the population, resulting in potential errors. Statistical hypothesis testing is the process used to determine if claims from samples to populations can be made and with what certainty.

The null hypothesis (Ho) is the claim that there is no real difference between two groups, while the alternative hypothesis (H1 of Ha) suggests that any difference is due to some non-random chance. The alternative hypothesis can be one-tailed of two-tailed, depending on whether it seeks to establish a difference of a change in one direction.

Two types of errors may occur when testing the null hypothesis: Type I and Type II errors. Type I error occurs when the null hypothesis is rejected when it is true, while Type II error occurs when the null hypothesis is accepted when it is false. The power of a study is the probability of correctly rejecting the null hypothesis when it is false, and it can be increased by increasing the sample size.

P-values provide information on statistical significance and help researchers decide if study results have occurred due to chance. The p-value is the probability of obtaining a result that is as large of larger when in reality there is no difference between two groups. The cutoff for the p-value is called the significance level (alpha level), typically set at 0.05. If the p-value is less than the cutoff, the null hypothesis is rejected, and if it is greater or equal to the cut off, the null hypothesis is not rejected. However, the p-value does not indicate clinical significance, which may be too small to be meaningful.

A study that has a greater sample size is considered to have higher power, meaning it is capable of detecting a significant difference of effect that is clinically relevant.

The Importance of Power in Statistical Analysis

Power is a crucial concept in statistical analysis as it helps researchers determine the number of participants needed in a study to detect a clinically significant difference of effect. It represents the probability of correctly rejecting the null hypothesis when it is false, which means avoiding a Type II error. Power values range from 0 to 1, with 0 indicating 0% and 1 indicating 100%. A power of 0.80 is generally considered the minimum acceptable level.

Several factors influence the power of a study, including sample size, effect size, and significance level. Larger sample sizes lead to more precise parameter estimations and increase the study’s ability to detect a significant effect. Effect size, which is determined at the beginning of a study, refers to the size of the difference between two means that leads to rejecting the null hypothesis. Finally, the significance level, also known as the alpha level, represents the probability of a Type I error. By considering these factors, researchers can optimize the power of their studies and increase the likelihood of detecting meaningful effects.

There are several reporting guidelines available for different types of research studies. The COREQ checklist, consisting of 32 items, is designed for reporting qualitative research that involves interviews and focus groups. The CONSORT Statement provides a 25-item checklist to aid in reporting randomized controlled trials (RCTs). For reporting the pooled findings of multiple studies, the QUOROM and PRISMA guidelines are useful. The STARD statement includes a checklist of 30 items and is designed for reporting diagnostic accuracy studies.

Epidemiology Key Terms

– Epidemic (Outbreak): A rise in disease cases above the anticipated level in a specific population during a particular time frame.
– Endemic: The regular of anticipated level of disease in a particular population.
– Pandemic: Epidemics that affect a significant number of individuals across multiple countries, regions, of continents.

The Importance of Power in Statistical Analysis

Power is a crucial concept in statistical analysis as it helps researchers determine the number of participants needed in a study to detect a clinically significant difference of effect. It represents the probability of correctly rejecting the null hypothesis when it is false, which means avoiding a Type II error. Power values range from 0 to 1, with 0 indicating 0% and 1 indicating 100%. A power of 0.80 is generally considered the minimum acceptable level.

Several factors influence the power of a study, including sample size, effect size, and significance level. Larger sample sizes lead to more precise parameter estimations and increase the study’s ability to detect a significant effect. Effect size, which is determined at the beginning of a study, refers to the size of the difference between two means that leads to rejecting the null hypothesis. Finally, the significance level, also known as the alpha level, represents the probability of a Type I error. By considering these factors, researchers can optimize the power of their studies and increase the likelihood of detecting meaningful effects.

– Treating 1000 patients with a new statin for five years prevented 50 MIs.
– The number needed to treat (NNT) to prevent one MI is 20 (1000/50).
– NNT provides information on treatment efficacy beyond statistical significance.
– Based on these data, treating as few as 20 patients over five years may prevent an infarct.
– Cost economic data can be calculated by factoring in drug costs and costs of treating and rehabilitating a patient with an MI.

Wilson and Junger Criteria for Screening

1. The condition should be an important public health problem.
2. There should be an acceptable treatment for patients with recognised disease.
3. Facilities for diagnosis and treatment should be available.
4. There should be a recognised latent of early symptomatic stage.
5. The natural history of the condition, including its development from latent to declared disease should be adequately understood.
6. There should be a suitable test of examination.
7. The test of examination should be acceptable to the population.
8. There should be agreed policy on whom to treat.
9. The cost of case-finding (including diagnosis and subsequent treatment of patients) should be economically balanced in relation to the possible expenditure as a whole.
10. Case-finding should be a continuous process and not a ‘once and for all’ project.

The Wilson and Junger criteria provide a framework for evaluating the suitability of a screening program for a particular condition. The criteria emphasize the importance of the condition as a public health problem, the availability of effective treatment, and the feasibility of diagnosis and treatment. Additionally, the criteria highlight the importance of understanding the natural history of the condition and the need for a suitable test of examination that is acceptable to the population. The criteria also stress the importance of having agreed policies on whom to treat and ensuring that the cost of case-finding is economically balanced. Finally, the criteria emphasize that case-finding should be a continuous process rather than a one-time project. By considering these criteria, public health officials can determine whether a screening program is appropriate for a particular condition and ensure that resources are used effectively.

Recall bias pertains to how a person’s illness status can influence their tendency to report past exposure to a risk factor. Confounding arises when an additional variable is associated with both an independent and dependent variable. Observer bias refers to the possibility that researchers’ cognitive biases may unconsciously impact the results of a study. Publication bias refers to the tendency for studies with positive results to be more likely to be published. Selection bias occurs when certain individuals of groups are overrepresented, leading to inadequate randomization.

The relative risk (RR) is calculated by dividing the exposure event rate (EER) by the control event rate (CER). In this case, the EER is 10 out of 100 (0.10) and the CER is 50 out of 300 (0.166). Therefore, the RR is calculated as 0.10 divided by 0.166, which equals 0.6.

Measures of Effect in Clinical Studies

When conducting clinical studies, we often want to know the effect of treatments of exposures on health outcomes. Measures of effect are used in randomized controlled trials (RCTs) and include the odds ratio (of), risk ratio (RR), risk difference (RD), and number needed to treat (NNT). Dichotomous (binary) outcome data are common in clinical trials, where the outcome for each participant is one of two possibilities, such as dead of alive, of clinical improvement of no improvement.

To understand the difference between of and RR, it’s important to know the difference between risks and odds. Risk is a proportion that describes the probability of a health outcome occurring, while odds is a ratio that compares the probability of an event occurring to the probability of it not occurring. Absolute risk is the basic risk, while risk difference is the difference between the absolute risk of an event in the intervention group and the absolute risk in the control group. Relative risk is the ratio of risk in the intervention group to the risk in the control group.

The number needed to treat (NNT) is the number of patients who need to be treated for one to benefit. Odds are calculated by dividing the number of times an event happens by the number of times it does not happen. The odds ratio is the odds of an outcome given a particular exposure versus the odds of an outcome in the absence of the exposure. It is commonly used in case-control studies and can also be used in cross-sectional and cohort study designs. An odds ratio of 1 indicates no difference in risk between the two groups, while an odds ratio >1 indicates an increased risk and an odds ratio <1 indicates a reduced risk.

The observed link between alcohol consumption and lung cancer is likely due to confounding factors, such as cigarette smoking. Confounding variables are those that are associated with both the independent and dependent variables, in this case, alcohol consumption and lung cancer.

Types of Bias in Statistics

Bias is a systematic error that can lead to incorrect conclusions. Confounding factors are variables that are associated with both the outcome and the exposure but have no causative role. Confounding can be addressed in the design and analysis stage of a study. The main method of controlling confounding in the analysis phase is stratification analysis. The main methods used in the design stage are matching, randomization, and restriction of participants.

There are two main types of bias: selection bias and information bias. Selection bias occurs when the selected sample is not a representative sample of the reference population. Disease spectrum bias, self-selection bias, participation bias, incidence-prevalence bias, exclusion bias, publication of dissemination bias, citation bias, and Berkson’s bias are all subtypes of selection bias. Information bias occurs when gathered information about exposure, outcome, of both is not correct and there was an error in measurement. Detection bias, recall bias, lead time bias, interviewer/observer bias, verification and work-up bias, Hawthorne effect, and ecological fallacy are all subtypes of information bias.

When comparing new diagnostic tests with gold standard tests, work-up bias can be a concern. Clinicians may be hesitant to order the gold standard test unless the new test yields a positive result, as the gold standard test may involve invasive procedures like tissue biopsy. This can significantly skew the study’s findings and affect metrics such as sensitivity and specificity. While it may not always be possible to eliminate work-up bias, researchers must account for it in their analysis.

Types of Bias in Statistics

Bias is a systematic error that can lead to incorrect conclusions. Confounding factors are variables that are associated with both the outcome and the exposure but have no causative role. Confounding can be addressed in the design and analysis stage of a study. The main method of controlling confounding in the analysis phase is stratification analysis. The main methods used in the design stage are matching, randomization, and restriction of participants.

There are two main types of bias: selection bias and information bias. Selection bias occurs when the selected sample is not a representative sample of the reference population. Disease spectrum bias, self-selection bias, participation bias, incidence-prevalence bias, exclusion bias, publication of dissemination bias, citation bias, and Berkson’s bias are all subtypes of selection bias. Information bias occurs when gathered information about exposure, outcome, of both is not correct and there was an error in measurement. Detection bias, recall bias, lead time bias, interviewer/observer bias, verification and work-up bias, Hawthorne effect, and ecological fallacy are all subtypes of information bias.

Logistic regression is employed when dealing with dichotomous variables, which are variables that have only two possible values, such as live/dead of head/tail.

Stats: Correlation and Regression

Correlation and regression are related but not interchangeable terms. Correlation is used to test for association between variables, while regression is used to predict values of dependent variables from independent variables. Correlation can be linear, non-linear, of non-existent, and can be strong, moderate, of weak. The strength of a linear relationship is measured by the correlation coefficient, which can be positive of negative and ranges from very weak to very strong. However, the interpretation of a correlation coefficient depends on the context and purposes. Correlation can suggest association but cannot prove of disprove causation. Linear regression, on the other hand, can be used to predict how much one variable changes when a second variable is changed. Scatter graphs are used in correlation and regression analyses to visually determine if variables are associated and to detect outliers. When constructing a scatter graph, the dependent variable is typically placed on the vertical axis and the independent variable on the horizontal axis.

Gender has traditionally been recorded as either male of female, creating a binary of dichotomous variable. Other categorical variables, such as eye color and ethnicity, can be grouped into two or more categories. Continuous variables, such as temperature, height, weight, and age, can be placed anywhere on a scale and have mathematical properties. Ordinal variables allow for ranking, but do not allow for direct mathematical comparisons between values.

Stratification is a method of managing confounding by dividing the data into two or more groups where the confounding variable remains constant of varies minimally.

Types of Bias in Statistics

Bias is a systematic error that can lead to incorrect conclusions. Confounding factors are variables that are associated with both the outcome and the exposure but have no causative role. Confounding can be addressed in the design and analysis stage of a study. The main method of controlling confounding in the analysis phase is stratification analysis. The main methods used in the design stage are matching, randomization, and restriction of participants.

There are two main types of bias: selection bias and information bias. Selection bias occurs when the selected sample is not a representative sample of the reference population. Disease spectrum bias, self-selection bias, participation bias, incidence-prevalence bias, exclusion bias, publication of dissemination bias, citation bias, and Berkson’s bias are all subtypes of selection bias. Information bias occurs when gathered information about exposure, outcome, of both is not correct and there was an error in measurement. Detection bias, recall bias, lead time bias, interviewer/observer bias, verification and work-up bias, Hawthorne effect, and ecological fallacy are all subtypes of information bias.

To address the issue of drop-outs in a study, an intention to treat (ITT) analysis can be employed. Drop-outs can lead to attrition bias, which creates systematic differences in attrition across treatment groups. In an ITT analysis, all patients are included in the groups they were initially assigned to through random allocation. To handle missing data, two common methods are last observation carried forward and worst case scenario analysis.

Disease Rates and Their Interpretation

Disease rates are a measure of the occurrence of a disease in a population. They are used to establish causation, monitor interventions, and measure the impact of exposure on disease rates. The attributable risk is the difference in the rate of disease between the exposed and unexposed groups. It tells us what proportion of deaths in the exposed group were due to the exposure. The relative risk is the risk of an event relative to exposure. It is calculated by dividing the rate of disease in the exposed group by the rate of disease in the unexposed group. A relative risk of 1 means there is no difference between the two groups. A relative risk of <1 means that the event is less likely to occur in the exposed group, while a relative risk of >1 means that the event is more likely to occur in the exposed group. The population attributable risk is the reduction in incidence that would be observed if the population were entirely unexposed. It can be calculated by multiplying the attributable risk by the prevalence of exposure in the population. The attributable proportion is the proportion of the disease that would be eliminated in a population if its disease rate were reduced to that of the unexposed group.

Snowball sampling is a unique technique utilized in qualitative research when the desired sample trait is uncommon. In such cases, it can be challenging of expensive to locate suitable respondents. Snowball sampling involves existing subjects recruiting future subjects, which can help overcome these difficulties. For more information on this method, please refer to the additional resources provided.

Case-studies and case-series can have an experimental nature due to the potential involvement of interventions of treatments.

Types of Primary Research Studies and Their Advantages and Disadvantages

Primary research studies can be categorized into six types based on the research question they aim to address. The best type of study for each question type is listed in the table below. There are two main types of study design: experimental and observational. Experimental studies involve an intervention, while observational studies do not. The advantages and disadvantages of each study type are summarized in the table below.

Type of Question Best Type of Study

Therapy Randomized controlled trial (RCT), cohort, case control, case series
Diagnosis Cohort studies with comparison to gold standard test
Prognosis Cohort studies, case control, case series
Etiology/Harm RCT, cohort studies, case control, case series
Prevention RCT, cohort studies, case control, case series
Cost Economic analysis

Study Type Advantages Disadvantages

Randomized Controlled Trial – Unbiased distribution of confounders – Blinding more likely – Randomization facilitates statistical analysis – Expensive – Time-consuming – Volunteer bias – Ethically problematic at times
Cohort Study – Ethically safe – Subjects can be matched – Can establish timing and directionality of events – Eligibility criteria and outcome assessments can be standardized – Administratively easier and cheaper than RCT – Controls may be difficult to identify – Exposure may be linked to a hidden confounder – Blinding is difficult – Randomization not present – For rare disease, large sample sizes of long follow-up necessary
Case-Control Study – Quick and cheap – Only feasible method for very rare disorders of those with long lag between exposure and outcome – Fewer subjects needed than cross-sectional studies – Reliance on recall of records to determine exposure status – Confounders – Selection of control groups is difficult – Potential bias: recall, selection
Cross-Sectional Survey – Cheap and simple – Ethically safe – Establishes association at most, not causality – Recall bias susceptibility – Confounders may be unequally distributed – Neyman bias – Group sizes may be unequal
Ecological Study – Cheap and simple – Ethically safe – Ecological fallacy (when relationships which exist for groups are assumed to also be true for individuals)

In conclusion, the choice of study type depends on the research question being addressed. Each study type has its own advantages and disadvantages, and researchers should carefully consider these when designing their studies.

To prevent bias in the study, the researchers are implementing patient blinding to prevent performance bias, as knowledge of whether they are taking venlafaxine of a placebo, of which arm of the study they are in, could impact the patient’s behavior. Additionally, investigators must also be blinded to avoid measurement bias.

SIGLE is a database that contains unpublished of ‘grey’ literature, while CINAHL is a database that focuses on healthcare and biomedical journal articles. The Cochrane Library is a collection of databases that includes the Cochrane Reviews, which are systematic reviews and meta-analyses of medical research. EMBASE is a pharmacological and biomedical database, and PsycINFO is a database of abstracts from psychological literature that is created by the American Psychological Association.

The Relative Risk Reduction (RRR) is calculated by subtracting the experimental event rate (EER) from the control event rate (CER), dividing the result by the CER, and then multiplying by 100 to get a percentage. In this case, the RRR is (35-20)÷35 = 0.4285 of 42.85%.