INTRODUCTIONThe aim of this study was to assess the sensitivity of McDonald's magnetic resonance imaging (MRI) criteria for the diagnosis of multiple sclerosis (MS) in a group of Asian patients diagnosed with clinically definite MS, based on lesion characterisation on MRI scans.

MATERIALS AND METHODSForty-nine patients from 3 major neurological institutions were classified as having Asian- or Western-type MS based on clinical assessment. Each MRI scan was reviewed by 2 neuroradiologists for the presence and characteristics of brain and spinal lesions. The McDonald's MRI criteria were then applied and its sensitivity evaluated.

RESULTSNine patients were excluded, leaving 34 females and 6 males who were dominantly Chinese (90%), with a mean age of 36.2 years. The MRI brain and spinal findings were detailed and tabulated. Statistically significant differences (P <0.01) in MRI brain findings and sensitivity of McDonald's MRI criteria were found between our Asian- and Western-type MS patients. The diagnostic yield of McDonald's MRI criteria increased by 20% when we substituted a cord for a brain lesion, and applied the substitution for enhancing cord lesions as well.

CONCLUSIONThe diagnosis is more likely to be made when using McDonald MRI criteria based on brain findings, in a patient who presents clinically with Western-type MS. The provision for substitution of "one brain for a spinal lesion" is helpful in Asian-type MS, where there is preponderance of spinal lesion load. Our findings suggest that minor modifications in the interpretation of McDonald's MRI criteria have significant impact on the diagnosis in patients clinically presenting as Asian-type MS, with potential bearing on their subsequent management.

Adult ; Brain Injuries ; diagnosis ; pathology ; Diagnosis, Differential ; Female ; Humans ; Magnetic Resonance Imaging ; Male ; Medical Audit ; Multiple Sclerosis ; classification ; diagnosis ; Retrospective Studies ; Sensitivity and Specificity ; Singapore ; Spinal Cord Injuries ; diagnosis ; pathology

Treatment-induced neuropathy of diabetes (TIND) is an acute painful autonomic small-fiber neuropathy that develops following an abrupt improvement in glycaemia control. Recent reports suggest TIND is a significant problem in tertiary neuropathy clinics. TIND in hospitalized patients with poor initial glycaemia control, that we refer to as nosocomial TIND, has not been well-studied. We describe the demographic, clinical features and indices of glycaemia control in 5 consecutive nosocomial TIND patients. TIND was defined using recently published criteria. Pre-meal capillary blood glucose recordings performed during the period of HbA1c decline was used to calculate glycaemic variability. All the nosocomial TIND patients were hospitalized for prolonged periods for serious medical conditions that warranted good glycaemia control, namely severe sepsis, diabetic ketoacidosis, stroke, heart failure and traumatic head injury. They had raised, double-digit, HbA1c levels at admission that subsequently dropped precipitously with tight in-patient glycaemia control protocols. These patients had multiple, largely asymptomatic, hypoglycaemic episodes. Glycaemic variability also appeared to be high in this cohort. TIND may be a significant cause of morbidity in hospitalized diabetic patients with poor glycaemia control. Not all patients developed both autonomic and painful neuropathies, raising the possibility of forme-fruste TIND