The reliability of calf and thigh measurements for the detection of deep vein thrombosis in patients with spinal cord injuries

Ellen Eugarde Bachelor of Nursing, Bachelor of Nursing Informatics

30/05/2013 School of Nursing and Midwifery Faculty of Health Griffith University

Submitted in fulfilment of the requirements of the degree of Master of Philosophy

Abstract Venous thrombo-embolism (VTE) has been recognised as a leading cause of morbidity and mortality in the acute spinal cord injury (SCI) population. It is thought that deep vein thrombosis (DVT), the most common form of VTE will occur in the majority of patients in the absence of adequate prophylaxis in the SCI group. Pulmonary embolism, which is often preceded by a DVT, is a leading cause of death in the first year after injury for this group of patients. The reliability of signs and symptoms in the diagnosis of thrombo-embolism has been debated in medicine since the 1960s; however diagnosis of DVT remains problematic. The most commonly used of these signs and symptoms in the SCI patient group is swelling of the calves and/or thighs as indicated by an increase in a daily circumference measurement. After a review of the literature it was identified that there were several aspects of reliability of the current practice of lower limb circumference measurements that had been poorly examined including the location the measurements are performed, the landmarks used to determine the location to be measured, the type of tape measure being used and the inter-rater reliability of the procedure. The aim of this study was to test the reliability of calf and thigh circumference measurements by testing the stability and equivalence of the equipment, the procedure and the measurements. This three phase prospective study was designed to determine how much the Birch tape measure stretches over time, if the standard (Birch) tape measure can be used instead of the Gulick spring tape for circumference measures, at which measurement location could the most consistent measurements be produced and whether it was possible for staff to consistently reproduce measurements over time.

i

In Phase I 20 Birch tape measures currently in clinical use were examined for stretch over time. Tape measures were found to have stretched between 1 and 7mm but this was not dependent on time in use. In Phase II the limits of agreement method was used to assess the equivalency of the Birch tape with the Gulick spring tape at the calf and thigh location. A total of 402 measurements at each location were collected. The limits of agreement for the measurement differences between the two tapes were wider at both locations (calf: -3.66 to 1.66 cm and thigh: -6.19 to 2.79 cm) than the +/- 1.5 cm set a priori. Some of this lack of agreement was attributed to poor reproducibility with both tape measures, with reproducibility coefficients wider than the limits of agreement at both the calf (Birch +/- 2.64cm and Gulick +/- 2.84 cm) and thigh (Birch +/- 4.15 cm and for the Gulick +/- 4.65 cm) locations. In Phase III the consistency of measurements performed at four separate locations, the ankle, mid-calf, calf and thigh across three time points was assessed. Relative standard deviations calculated from the first time point (baseline) indicated that calf measurements were the most consistently performed followed by the ankle measurements. Mid-calf and thigh measurements were found to be the least consistent. Two way repeated measures analysis of variance were performed for all four locations to assess inter-rater reliability. No significant variation in measurement was found between examiners at any of the locations. The effect of time by group membership was found to be approaching significance at the mid-calf level only (p=0.05). A significant effect of time on measurements was found at the mid-calf level (p=0.002) and the calf level (p=