Harnessing historical data to derive reference limits - A comparison of e-norms to traditionally derived reference limits

Objective

Nerve conduction studies (NCS) require valid reference limits for meaningful interpretation. We aimed to further develop the extrapolated norms (e-norms) method for obtaining NCS reference limits from historical laboratory datasets for children and adults, and to validate it against traditionally derived reference limits.

Methods

We compared reference limits obtained by applying a further developed e-norms with reference limits from healthy controls for the age strata's 9-18, 20-44 and 45-60 years old. The control data consisted of 65 healthy children and 578 healthy adults, matched with 1294 and 5628 patients respectively. Five commonly investigated nerves were chosen: The tibial and peroneal motor nerves (amplitudes, conduction velocities, F-waves), and the sural, superficial peroneal and medial plantar sensory nerves (amplitudes, conduction velocities). The datasets were matched by hospital to ensure identical equipment and protocols. The e-norms method was adapted, and reference limit calculation using both ±2 SD (original method) and ±2.5 SD (to compensate for predicted underestimation of population SD by the e-norms method) was compared to control data using ±2 SD. Percentage agreement between e-norms and the traditional method was calculated.

Results

On average, the e-norms method (mean ±2 SD) produced slightly stricter reference limits compared to the traditional method. Increasing the e-norms range to mean ±2.5 SD improved the results in children while slightly overcorrecting in the adult group. The average agreement between the two methods was 95 % (±2 SD) and 96 % (±2.5 SD).

Conclusions

The e-norms method yielded slightly stricter reference limits overall than ones obtained through traditional methods; However, much of the difference can be attributed to a few outlying plots where the raters found it difficult to apply e-norms correctly. The two methods disagreed on classification of 4-5% of cases. Our e-norms software is suited to analyze large amounts of raw NCS data; it should further reduce bias and facilitate more accurate ratings.

Significance

With small adaptations, the e-norms method adequately replicates traditionally derived reference limits, and is a viable method to produce reference limits from historical datasets.

Diagnostic accuracy of the 5.07 monofilament test for diabetes polyneuropathy: influence of age, sex, neuropathic pain and neuropathy severity

INTRODUCTION

There is a need for simple and cheap diagnostic tools for diabetic polyneuropathy (DPN). We aimed to assess the diagnostic accuracy of the 5.07/10 g monofilament test in patients referred to polyneuropathy assessments, as well as to examine how disease severity, age, sex and neuropathic pain (NP) impact diagnostic accuracy.

RESEARCH DESIGN AND METHODS

Five Norwegian university hospitals recruited patients with diabetes aged 18-70 referred to neurological outpatient clinics for polyneuropathy assessments. The 5.07/10 g Semmes-Weinstein monofilament examination (SWME) was validated against the Toronto consensus for diagnosing diabetic neuropathies; the results were stratified by age, sex and NP. Disease severity was graded by a combined nerve conduction study (NCS) Z-score, and logistic regression was applied to assess whether disease severity was a predictor of diagnostic accuracy.

RESULTS

In total, 506 patients were included in the study. Global sensitivity was 0.60 (95% CI 0.55, 0.66), specificity 0.82 (95% CI 0.75, 0.87), positive and negative predictive values were 0.86 (95% CI 0.81, 0.90) and 0.52 (95% CI 0.46, 0.58), respectively, positive and negative likelihood ratios were 3.28 (95% CI 2.37, 4.53) and 0.49 (95% CI 0.42, 0.57), respectively. The SWME was less sensitive in females (0.43), had lower specificity in patients with NP (0.56), and performed worse in patients ≥50 years. NCS-based disease severity did not affect diagnostic accuracy (OR 1.15, 95% CI 0.95, 1.40).

CONCLUSIONS

This multicenter study demonstrates poor diagnostic performance for the 5.07/10 g SWME in patients with diabetes referred to polyneuropathy assessments; it is particularly unsuited for female patients and those with NP. The diagnostic accuracy of the SWME was not influenced by NCS-based disease severity, demonstrating that it does not perform better in patients with later stages of DPN. We do not recommend the use of the 5.07/10 g monofilament in the evaluation of patients with diabetes referred to polyneuropathy assessments.

Estimating motor unit numbers from a CMAP scan: Repeatability study on three muscles at 15 centres

OBJECTIVE

To assess the repeatability and suitability for multicentre studies of MScanFit motor unit number estimation (MUNE), which involves modelling compound muscle action potential (CMAP) scans.

METHODS

Fifteen groups in 9 countries recorded CMAP scans twice, 1-2 weeks apart in healthy subjects from abductor pollicis brevis (APB), abductor digiti minimi (ADM) and tibialis anterior (TA) muscles. The original MScanFit program (MScanFit-1) was compared with a revised version (MScanFit-2), designed to accommodate different muscles and recording conditions by setting the minimal motor unit size as a function of maximum CMAP.

RESULTS

Complete sets of 6 recordings were obtained from 148 subjects. CMAP amplitudes differed significantly between centres for all muscles, and the same was true for MScanFit-1 MUNE. With MScanFit-2, MUNE differed less between centres but remained significantly different for APB. Coefficients of variation between repeats were 18.0% for ADM, 16.8% for APB, and 12.1% for TA.

CONCLUSIONS

It is recommended for multicentre studies to use MScanFit-2 for analysis. TA provided the least variable MUNE values between subjects and the most repeatable within subjects.

SIGNIFICANCE

MScanFit was primarily devised to model the discontinuities in CMAP scans in patients and is less suitable for healthy subjects with smooth scans.

Which combined nerve conduction study scores are best suited for polyneuropathy in diabetic patients?

INTRODUCTION/AIMS

Nerve conduction studies (NCS) are widely used in diagnosing diabetic polyneuropathy. Combining the Z scores of several measures (Z-compounds) may improve diagnostics by grading abnormality. We aimed to determine which combination of nerves and measures is best suited for studies of diabetic polyneuropathy.

METHODS

Sixty-eight patients with type 1 diabetes and 35 controls were included in this study. NCS measurements were taken from commonly investigated nerves in one arm and both legs. Different Z-compounds were calculated and compared with reference material to assess abnormality. A sensitivity proxy, the accuracy index (AI), and Cohen's d were calculated.

RESULTS

Z-compounds with the highest AI consisted of the tibial and peroneal motor, and the sural, superficial peroneal, and tibial medial plantar sensory nerves in one or two legs. All Z-compounds were able to discriminate between diabetic subjects and nondiabetic controls (mean Cohen's d = 1.42 [range, 1.03-1.63]). The association between AI and number of measures was best explained logarithmically (R²  = 0.401), with diminishing returns above approximately 14 or 15 measures. F-wave inclusion may increase the AI of the Z compounds. Although often clinically useful among the non-elderly, the additional inclusion of medial plantar NCS into Z-compounds in general did not improve AI.

DISCUSSION

Performing unilateral NCS in several motor and sensory lower extremity nerves is suited for the evaluation of polyneuropathy in diabetic patients. The use of Z-compounds may improve diagnostic accuracy in diabetic polyneuropathy and may be particularly useful for follow-up research studies as single summary measures of NCS abnormality development over time.

Can within-subject comparisons of thermal thresholds be used for diagnostic purposes?

•

Normal limits for within-subject comparisons of thermal thresholds are wide.

•

Our findings advocate for site-specific normal values of adequate resolution.

•

The difference between distal and proximal thresholds increase drastically with age.

Objective

Quantitative thermal testing (QTT) is a psychophysical assessment method of small nerve fibers that relies on reference material to assess function. Normal limits for within-subject comparisons of thermal thresholds are scarce, and their association with age, height and sex is not fully elucidated. The aim of this study was to investigate the normal limits for distal-proximal– and contralateral homologous comparisons of thermal thresholds with QTT, and their association with age, sex or height.

Methods

Fifty healthy volunteers ages 20–79 participated in the experiment. Cold detection thresholds (CDT), warm detection thresholds (WDT), heat pain thresholds (HPT), and cold pain thresholds (CPT) were measured bilaterally at the thenar eminence, anterior thigh, distal medial leg and foot dorsum. Sample normal limits were calculated as (mean) ± 2 SD.

Results

Forty-eight subjects were included in the analysis. CPT was excluded from all analyses due to a large floor-effect. Sample normal limits for side-differences ranged from 1.8 to 7.2 °C for CDT, 2.4–6.8 °C for WDT and 3.2–4.0 °C for HPT, depending on anatomical site. For distal-proximal comparisons, sample normal limits ranged from 4.0 to 8.7 °C for CDT, 6.0–14.0 °C for WDT and 4.2–9.0 °C for HPT, depending on the pairs compared. Age was associated with side-differences for CDT in the thenar eminences (p < 0.001) and distal medial legs (p < 0.002), and with 11 of 18 distal-proximal comparisons (p < 0.01).

Conclusions

The normal limits for distal-proximal- and contralateral homologous thermal thresholds were wide, and thus of limited use in a clinical setting, although the reported values may be somewhat inflated by low sample-size and consequent age-pooling. Age, but not sex or height, was associated with contralateral differences in CDT in the thenar eminences and distal medial legs, and with most distal-proximal differences.

Significance

Due to wide normal limits, we advise caution when utilizing relative comparisons of thermal thresholds for diagnostic purposes.