Solving Statistics: Stepped-wedge randomized controlled trial

Stepped-wedge randomized controlled trial

Background

The performance of activities of daily living (ADL) at home is important for the recovery of older individuals after hip fracture. However, 20-90% of these individuals lose ADL function and never fully recover. Although exercise interventions have been proven to improve physical function, especially elderly do not seem to benefit from these interventions. In this prospective, stepped-wedge randomized controlled trial, care as usual [CaU] is compared to 1) occupational therapy (OT) with coaching based on cognitive behavioural treatment (CBT) [OTc], and 2) OT-CBT with sensor monitoring embedded [OTcsm]. More specifically, during 12 months, six nursing homes will start with providing CaU, then cross over to provide OTc and finally cross over to provide OTcsm. The timing of crossing over is randomized: two nursing homes will cross over for the first time after two months, two after four months and the last two after six months. OTc will always be provided for 4 months, CaU for two, four and siex months respectively, and OTcsm for six, four and two months. The primary outcome measure, perceived daily functioning, is measured 6 months after start of rehabilitation and compared to baseline functioning.1

Solving Statistics: Should I use Pearson’s or Spearman’s correlation coefficient

Should I correct for multiple testing and what is the Bonferroni correction?

Background

A correlation coefficient is a number that shows if there is an association between variables. It provides an indication of the association between two variables X and Y (1). Two types of correlations are commonly used for continuous, numerical data. These are Pearson’s correlation coefficient and Spearman’s rank correlation coefficient.

Solving Statistics: p-values and confidence intervals

p-values and confidence intervals

No statistical question this time. Partly because I wasn’t asked a question, but also because of a recent Science Cafe organised by the Young Statisticians of the Netherlands Society for Statistics and Operations Research. The subject of this evening was ‘To p or not to p?’, referring to the widely reported as well as misinterpreted and even mistrusted p-value (medical) researchers love to report. Unfortunately I was unable to attend, it would have been an interesting topic. A discussion with medical students at the VUmc during a statistics course made me wonder: “Shouldn’t we actually start with explaining basic statistical terminology before we can start performing analyses?”. Therefore, this ‘statistical problem’ is about p-values (and a little bit about confidence intervals).

Solving Statistics: Should I correct for multiple testing and what is the Bonferroni correction?

Should I correct for multiple testing and what is the Bonferroni correction?

Background

A hypothesis test is a method of statistical inference on sets of random variables, such as Hand Eczema Severity Index (HECSI) scores obtained from patients with hand eczema following treatment with hand creams “handy help” or “silky smooth”. If the values of the HECSI are normally distributed, a researcher can use Student’s t-test to compare the mean HECSI scores in both groups. The null hypothesis is that the mean scores are equal in both groups and the alternative hypothesis that the mean scores are not equal.

Solving Statistics: A statistical problem – use of parametric tests for small samples?

A statistical problem – use of parametric tests for small samples?

Question

As a junior researcher I noticed that there are different opinions on when to choose nonparametric tests (like the Mann-Whitney or Kruskal-Wllis test) over parametric tests (like the independent samples t-test or ANOVA). Most researchers know that this decision should be made based on the distribution of the data: parametric tests for normally distributed outcomes, nonparametric tests for non-normal data. Therefore, in every beginners course on Statistics different ways to test/assess normality are discussed (histograms, QQ-plots, the Kolgomorov-Smirnov test, and the Shapiro-Wilk test).

Solving Statistics: Should I test for differences in baseline characteristics, for example in a randomised controlled study?

Should I test for differences in baseline characteristics, for example in a randomised controlled study?

Background

Researchers usually present the characteristics of the participants in each group at the start of a study in a table. This table is often the first table in a paper and, hence, called Table 1. This table gives the reader an overview of the study participants and examines whether the participants are similar to patients he or she encounters. The reader can also use the information in the table to judge whether the participants in the two groups were comparable. Sometimes the two groups differ with respect to relevant demographic and clinical characteristics. Then it is important to correct for these differences in further analyses and take them into account when interpreting the results of the study.

Question

I am analyzing data from a small randomized controlled clinical trial with two arms. Should I test whether there are differences in baseline characteristics between the two groups and present the p-values in Table 1 of my manuscript? What do the results of these tests mean for further analysis that I carry out?

Case Report: An unexpected diagnosis in a newborn with severe prolonged hyperbilirubinemia without hemolysis

An unexpected diagnosis in a newborn with severe prolonged hyperbilirubinemia without hemolysis

This case describes a full term baby boy of a healthy mother born after a normal pregnancy, who developed jaundice on the second day after birth, with a total serum bilirubin (TSB) of 25.0 mg/dL (427 μmol/L) on the ninth day. Initial blood tests excluded common causes of neonatal hyperbilirubinemia, e.g. iso-immunization disorders, hemolysis and hypothyroidism. As bilirubin levels continued to remain high despite phototherapy, further investigation was warranted, revealing a decreased glucose 6 phosphate dehydrogenase (G6PD) activity in the red blood cells of the newborn. After eleven days of phototherapy the patient was discharged with a TSB of 21.5 mg/dL (368 μmol/L).

Background

Neonatal jaundice is commonly observed among newborn infants, caused by hyperbilirubinemia. Severe hyperbilirubinemia should be recognized and treated to prevent kernicterus, a condition characterized by irreversible neurological damage. In most cases, hyperbilirubinemia results from a physiological increase in the unconjugated bilirubin concentration, combined with immature mechanisms for conjugation and enhanced enterohepatic circulation. However, certain conditions (e.g. prolonged jaundice, onset in the first 24 hours after birth, rapid rise in serum bilirubin, etc.) should raise the suspicion of an underlying pathologic mechanism. In general, unconjugated hyperbilirubinemia can be caused by (1) an increased, pathologic production of bilirubin, (2) a deficiency of hepatic uptake, (3) an impaired conjugation of bilirubin, (4) an increased enterohepatic circulation of bilirubin, or (5) a combination of the above [1]. In case of pathologic unconjugated hyperbilirubinemia, an increased production of bilirubin due to hemolysis is the most likely cause. Therefore, a common approach in the diagnostic work-up of neonatal unconjugated hyperbilirubinemia is to differentiate between hemolytic and non-hemolytic diseases as a first step [2].

Article: Differences in trust in physician under 50 years between native and non-native patients: a single Dutch institute experience of 170 patients

Differences in trust in physician under 50 years between native and non-native patients: a single Dutch institute experience of 170 patients

Background Literature suggests that a patient’s ethnicity influences the degree of trust a patient has in his physician. This is of major importance, as trust influences health seeking behavior. No studies on ethnic differences in trust in physicians have been conducted yet in the Netherlands.
Objective To compare trust in physicians between native and non-native patients.
Methods To examine trust, we handed out questionnaires from December 2012 until April 2013, at the outpatient clinic of Internal Medicine at the AMC in Amsterdam. The questionnaires included the 10-item validated Wake Forest trust in physician scale. We examined both global trust and three different aspects: fidelity, competence and honesty.
Results One hundred seventy patients were included in the analysis of which 111 native Dutch and 59 non-native. Natives rated trust in their physicians on average 1.1 points (on a 10 point scale) higher (P=0.002). Especially natives in the age category of 18-50 years had more trust in their physicians than non-natives (P=0.002). However, this difference could not be found in the age category of >50 years. Furthermore, non-natives rated trust in their physician less often as ‘sufficient’ (at least 7 out of 10 points) (OR 0.37, 95% CI [0.16 – 0.88]). This especially accounted for the group 18-50 years (OR 0.19, 95% CI [0.045 – 0.76]). Of the three different aspects of trust, perceived physician competence was the strongest driver of ethnic differences in physician trust (OR 0.27, P<0.001, 95% CI [-0.33 – -0.10]).
Conclusion Native patients of 18-50 years show more trust in their physicians than non-natives. The knowledge obtained in this study should become a basis of a new strategy to improve physician trust in non-native patients.

Introduction

Cultural differences are known to cause a gap among citizens in general, but also between non-natives and their physicians1. This could affect the quality of healthcare these patients receive.

A premise for any patient-physician relationship is trust. The most important predictors of trust are similar to the predictors of patient satisfaction2. Furthermore, several studies have shown that patients who have more trust in their doctors show better therapy compliance3,4,5­. Also, research has shown that patients are more satisfied with female than male physicians6.

Solving Statistics: Inter-observer reliability

Inter-observer reliability

Background

In a prospective cohort study to propose a novel ultrasound scoring system for hand and wrist joints (US10) for evaluation of patients with early rheumatoid arthritis (RA), the researchers also investigated inter-observer reliability between two readers. Images from 20 randomly chosen patients (of the 48 patients included in the study) were evaluated for the inflammation and joint damage parameters of the joints included in the US10 score (second and third MCPs and PIPs and wrists), yielding a total of 200 images. A trained rheumatologist preformed the initial US examination, he had eight years of experience in US and was blinded to all other study findings. The 2nd evaluation of the ultrasound images was performed by a rheumatologist, with five years of experie

Radiology Image: An athlete with pain

An athlete with pain