Roaming through methodology. XXV. Outcome measures, surrogate outcomes, and intermediate measures

Clinical medicine is aimed at decreasing the current or future burden of disease. Disease outcome is best expressed as burden of disease or change thereof, and as such the only true 'outcome measure' in clinical research. However, outcome research is expensive in costs and time expenditure, and sometimes impossible. Therefore study results are often expressed in intermediate measures. These measures tell us something about the disease process and the pathophysiological consequences of the disease and should have a relation with outcome. If this relation is strong, the measure is called 'surrogate outcome'. Intermediate measures and surrogate outcomes have advantages and disadvantages. The reader of clinical trial results first has to decide whether the answer to the study question is relevant in his personal situation. If so, the applicability of a measure can be simply appraised by answering 3 questions: 'Is the measure truthful (relevant, unbiased)?'; 'Does it discriminate between situations that are of interest?'; 'Is the measure feasible in my setting?'     

The case for mild hyperhomocysteinaemia as a risk factor

The high incidence of vascular complications in severe hyperhomocysteinaemia in homozygotes for cystathionine -synthase deficiency has focused attention upon homocysteine as an atherogenic and thrombophilic agent. For two decades there has been accumulating evidence of mild hyperhomocysteinaemia as risk factor of vascular disease. Pooled data on hundreds of coronary, cerebrovascular and peripheral arterial disease patients show that mild hyperhomocysteinaemia was detectable in about 20-30%. In a recent meta-analysis of 27 studies up to 1994, including about 4000 patients and as many controls, it is calculated that the summary odds ratio of elevated homocysteine levels was 1.7, with 95% confidence interval (CI) 1.5-1.9, for coronary heart disease; it was 2.5, with 95% CI 2.0-3.0, for cerebro-vascular disease; and it was 6.8, with 95% CI 2.9-15.8, for peripheral vascular disease. The relevance of this newly recognized risk factor will be demonstrated by the outcome of the European Comac study on Hyperhomocysteinaemia and Vascular Disease, a multicentre case-control study on 800 vascular patients and 750 controls. Despite the selection for epidemiological reasons of a relatively low cut-off level as the criterion for mild hyperhomocysteinaemia in this study - the upper 20% of the distribution of control levels - the relative risk of thus-defined hyperhomocysteinaemia for arterial disease is about 2. This equals the relative risk of hypercholesterolaemia and of smoking; hypertension leads to a higher excess risk. The observed synergistic interaction between hyperhomocysteinaemia and hypertension and smoking may warrant a change in the now generally followed procedure of screening for hyperhomocysteinaemia only if conventional risk factors have not been detected in the patient. Those vascular patients with combined risk factors leading to synergism in their joint effect may profit most from homocysteine-lowering intervention.     

Polymorphisms in the transcobalamin gene: association with plasma homocysteine in healthy individuals and vascular disease patients

BACKGROUND: Hyperhomocysteinemia is an independent risk factor for cardiovascular disease (CVD). Intracellular vitamin B(12) deficiency may lead to increased plasma total homocysteine (tHcy) concentrations and because transcobalamin (TC) is the plasma transporter that delivers vitamin B(12) to cells, genetic variation in the TC gene may affect intracellular vitamin B(12) availability and, consequently, tHcy concentrations. METHODS: We examined five sequence variants, i.e., I23V, G94S, P259R, S348F, and R399Q, in the TC gene as possible determinants of tHcy and, concordantly, as possible risk factors for CVD in 190 vascular disease patients and 601 controls. We also studied potential effect-modification of vitamin B(12) by genotype. RESULTS: In individuals with high vitamin B(12), 259PP individuals had lower tHcy concentrations than 259PR and 259RR individuals. Homozygous 23VV individuals had lower fasting tHcy concentrations than their 23IV and 23II peers. None of the genotypes defined by the three other sequence variants showed an association with tHcy concentrations, nor was any TC genotype associated with an increased CVD risk. CONCLUSIONS: In individuals in the highest quartile of the vitamin B(12) distribution (>299 pmol/L), tHcy concentrations are lower in 259PP homozygotes than in 259PR and 259RR individuals. Therefore, 259PP individuals, who represent >25% of the general population, may be more susceptible to reduction of plasma tHcy concentrations by increasing the vitamin B(12) status.     

A new graph and scoring system simplified analysis of changing states: disease remissions in a rheumatoid arthritis clinical trial

BackgroundIn the setting of multiple remission and relapse periods of a chronic disease, simple endpoint analysis does not fully capture all relevant information, and we need methods to additionally describe both the duration of remission as well as the interruptions in this desired state. Probably the two-state continuous Markov process model comprises the best mathematical approach to data analysis. However, this approach is complex and not intuitive to clinicians. In this paper we propose a simple scoring system and a graph that can enhance the information about the remission experience in a trial or cohort study.MethodsThe continuity rewarded (‘ConRew’) score sums up periods in remission, and rewards extended periods by placing more value on uninterrupted periods than on interrupted periods. The ‘patient vector graph’ attempts to plot each patient's remission experience over time as a horizontal line (the ‘vector’) that is visible when the patient is in remission, but interrupted whenever relapse occurs. In this way a pattern is formed that conveys the number of patients experiencing remission, their individual total duration and interruptions, and time when these occur.ResultsIn a dataset of a randomized trial in early rheumatoid arthritis, the graph clearly showed both early and late benefit of one group over the other. The scoring system demonstrated the main benefit was in the number of remission periods, not in their ‘uninterruptedness’.ConclusionBoth approaches proved feasible and added extra information.     

Characterization of the nasopharyngeal and middle ear microbiota in gastroesophageal reflux-prone versus gastroesophageal reflux non-prone children

Otitis media (OM) is one of the most common pediatric infections worldwide, but the complex microbiology associated with OM is poorly understood. Previous studies have shown an association between OM and gastroesophageal reflux (GER) in children. Therefore, in order to bridge the gap in our current understanding of the interaction between GER and OM, we investigated the nasopharyngeal and middle ear microbiota of children suffering from GER-associated OM and OM only, using culture-independent 16S rRNA gene sequencing. Middle ear fluid, nasopharyngeal swabs, and clinical data were collected as part of a prospective pilot study conducted at the Department of Otorhinolaryngology of the Erasmus MC-Sophia Children’s Hospital, Rotterdam, the Netherlands. A total of 30 children up to 12 years of age who suffered from recurrent acute otitis media (AOM) (5), chronic otitis media with effusion (OME) (23), or both (2), and who were listed for tympanostomy tube placement, were included in the study. Nine children were included in the GER-associated OM cohort and 21 in the OM-only cohort. We found no obvious effect of GER on the nasopharyngeal and middle ear microbiota between the two groups of children. However, our results highlight the need to assess the true role of Alloiococcus spp. and Turicella spp. in children presenting with a high prevalence of recurrent AOM and chronic OME.