Hyperglycaemia, diabetes and risk of fragility fractures: observational and Mendelian randomisation studies

AIMS/HYPOTHESIS

Fragility fractures may be a complication of diabetes, partly caused by chronic hyperglycaemia. We hypothesised that: (1) individuals with hyperglycaemia and diabetes have increased risk of fragility fracture; (2) hyperglycaemia is causally associated with increased risk of fragility fracture; and (3) diabetes and fragility fracture jointly associate with the highest risk of all-cause mortality.

METHODS

In total, 117,054 individuals from the Copenhagen City Heart Study and the Copenhagen General Population Study (the Copenhagen studies) and 390,374 individuals from UK Biobank were included for observational and one-sample Mendelian randomisation (MR) analyses. Fragility fractures were defined as fractures at the hip, spine and arm (humerus/wrist), collected from national health registries. Summary data for fasting glucose and HbA1c concentrations from 196,743 individuals in the Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC) were combined with data on fragility fractures from the Copenhagen studies in two-sample MR analyses.

RESULTS

Higher fasting and non-fasting glucose and HbA1c concentrations were associated with higher risk of any fragility fracture (p<0.001). Individuals with vs without diabetes had HRs for fragility fracture of 1.50 (95% CI 1.19, 1.88) in type 1 diabetes and 1.22 (1.13, 1.32) in type 2 diabetes. One-sample MR supported a causal association between high non-fasting glucose concentrations and increased risk of arm fracture in the Copenhagen studies and UK Biobank combined (RR 1.41 [1.11, 1.79], p=0.004), with similar results for fasting glucose and HbA1c in two-sample MR analyses (ORs 1.50 [1.03, 2.18], p=0.03; and 2.79 [1.12, 6.93], p=0.03, respectively). The corresponding MR estimates for any fragility fracture were 1.18 (1.00, 1.41), p=0.06; 1.36 (0.89, 2.09), p=0.15; and 2.47 (0.95, 6.43), p=0.06, respectively. At age 80 years, cumulative death was 27% in individuals with fragility fracture only, 54% in those with diabetes only, 67% in individuals with both conditions and 17% in those with neither.

CONCLUSIONS/INTERPRETATION

Hyperglycaemia and diabetes are risk factors for fragility fracture and one- and two-sample MR analyses supported a causal effect of hyperglycaemia on arm fractures. Diabetes and previous fragility fracture jointly conferred the highest risk of death in the general population.

Genetic variation in solute carrier family 5 member 2 mimicking sodium-glucose co-transporter 2-inhibition and risk of cardiovascular disease and all-cause mortality: reduced risk not explained by lower plasma glucose

AIMS

Treatment with sodium-glucose co-transporter 2 (SGLT2)-inhibitors reduces the risk of cardiovascular disease and mortality, but the mechanism is unclear. We hypothesized that a functional genetic variant in solute carrier family 5 member 2 (SLC5A2), known to be associated with familial renal glucosuria, would mimic pharmacological SGLT2-inhibition, and thus provide an opportunity to examine potential mediators of the effects on lower risk of cardiovascular disease and mortality.

METHODS AND RESULTS

We examined 112 712 individuals from the Copenhagen City Heart Study and Copenhagen General Population Study (CCHS + CGPS), 488 687 from the UK Biobank, and 342 499 from FinnGen, genotyped for SLC5A2 rs61742739, c.1961A > G; p.(Asn654Ser). The 2.0% heterozygotes and 0.01% homozygotes were pooled as carriers and compared with the 98% non-carriers. First, we examined the risk of cardiovascular disease and mortality; second, whether carrying the variant was associated with potential mediators of the effect; and third, whether identified potential mediators could explain the observed reduced risk of cardiovascular disease and mortality. In the CCHS + CGPS, carriers vs. non-carries had a 31% lower risk of heart failure, 21% lower risk of myocardial infarction, 16% lower risk of ischaemic heart disease, and 22% lower risk of all-cause mortality. Corresponding values in meta-analyses of the three studies combined were lower risk by 10%, 6%, 6%, and 10%, respectively. The SLC5A2 rs61742739 variant was not associated with a risk of ischaemic stroke or cardiovascular mortality. Of the lower risks observed in CCHS + CGPS, lower plasma glucose mediated 2.0%(P = 0.004) on heart failure, 3.1%(P = 0.09) on myocardial infarction, 4.1%(P = 0.02) on ischaemic heart disease, and 6.0%(P = 0.39) on all-cause mortality; corresponding values in the UK Biobank were 2.9%(P = 0.70), 1.5%(P = 0.77), 4.1%(P = 0.23), and 3.1%(P = 0.21), respectively.

CONCLUSION

A functional genetic variant in SLC5A2, mimicking SGLT2-inhibition, was associated with a lower risk of heart failure, myocardial infarction, ischaemic heart disease, and all-cause mortality. These effects were at most minimally mediated through lower plasma glucose.

Can a modified ketogenic diet be a nutritional strategy for patients with McArdle disease? Results from a randomized, single-blind, placebo-controlled, cross-over study

BACKGROUND

McArdle disease is caused by myophosphorylase deficiency leading to blocked glycogenolysis in skeletal muscle. Consequently, individuals with McArdle disease have intolerance to physical activity, muscle fatigue, and pain. These symptoms vary according to the availability of alternative fuels for muscle contraction. In theory, a modified ketogenic diet (mKD) can provide alternative fuels in the form of ketone bodies and potentially boost fat oxidation.

METHODS

This randomized, single-blind, placebo-controlled, cross-over study aimed to investigate if a mKD improves exercise capacity in individuals with McArdle disease. Participants were randomized to follow a mKD (75-80% fat, 15% protein, 5-10% carbohydrates) or placebo diet (PD) first for three weeks, followed by a wash-out period, and then the opposite diet. The primary outcome was change in heart rate during constant-load cycling. Secondary outcomes included change in plasma metabolites, perceived exertion, indirect calorimetry measures, maximal exercise capacity, and patient-reported outcomes.

RESULTS

Fifteen out of 20 patients with genetically verified McArdle disease completed all study visits, and 14 were included in the data analyses. We found that the mKD induced a metabolic shift towards increased fat oxidation (∼60% increase), and a 19-fold increase in plasma β-hydroxybutyrate (p < 0.05). The mKD did not improve heart rate responses during constant-load cycling but did improve patient-reported outcomes and maximal exercise capacity (∼20% increase) compared to the PD.

CONCLUSION

The mKD did not alleviate all McArdle disease-related symptoms but did induce some positive changes. To date, no satisfactory treatment options exist other than exercise training. To that end, a mKD can be a possible nutritional strategy for some individuals with McArdle disease who are motivated to undertake a restrictive diet.

CLINICAL TRIAL REGISTRATION

clinical trials.gov: NCT04044508.

SGLT2-inhibition increases total, LDL, and HDL cholesterol and lowers triglycerides: Meta-analyses of 60 randomized trials, overall and by dose, ethnicity, and drug type

BACKGROUND AND AIMS

Sodium glucose co-transporter 2 (SGLT2)-inhibitors were developed as glucose-lowering drugs. Surprisingly, SGLT2-inhibitors also reduced risk of cardiovascular disease. The impact of SGLT2-inhibitors on lipids and lipoproteins is unclear, but an effect might contribute to the observed lower cardiovascular risk. We conducted a meta-analysis to examine this, overall and by dose, ethnicity, and drug type.

METHODS

PubMed, EMBASE and Web of Science were searched for randomized controlled trials examining all available SGLT2-inhibitors. Studies with available lipid measurements were included. Quantitative data synthesis was performed using random and fixed effects models.

RESULTS

We identified 60 randomized trials, including 147,130 individuals. Overall, using random effects models, SGLT2-inhibitor treatment increased total cholesterol by 0.09 mmol/L (95% CI: 0.06, 0.13), low-density lipoprotein (LDL) cholesterol by 0.08 mmol/L (0.05, 0.10), and high-density lipoprotein (HDL) cholesterol by 0.06 mmol/L (0.05, 0.07), while it reduced triglycerides by 0.10 mmol/L (0.06, 0.14). Fixed effects estimates were similar but with smaller effect sizes for HDL cholesterol and triglycerides. For higher SGLT2-inhibitor doses, there was a nominally higher non-significant effect on lipids and lipoproteins. In Asian compared to non-Asian populations, a slightly larger increase in HDL cholesterol and a decrease in triglycerides were observed, but with similar results for total and LDL cholesterol. Treatment effects on lipids and lipoproteins were generally robust across different SGLT2-inhibitor drugs.

CONCLUSION

In meta-analyses, SGLT2-inhibition increased total, LDL, and HDL cholesterol and decreased triglycerides. Effect sizes varied slightly by drug dose and ethnicity but were generally robust by drug type.

Familial Hypercholesterolemia Prevalence Among Ethnicities—Systematic Review and Meta-Analysis

Background: Heterozygous familial hypercholesterolemia (FH) is a common genetic disorder leading to premature cardiovascular disease and death as a result of lifelong high plasma low-density lipoprotein cholesterol levels, if not treated early in life. The prevalence of FH varies between countries because of founder effects, use of different diagnostic criteria, and screening strategies. However, little is known about differences in FH prevalence according to ethnicity. We aimed to investigate the ethnic distribution of FH in diverse populations and estimate the prevalence of FH according to ethnicity.

Methods: We performed a systematic review and meta-analysis, searching PubMed and Web of Science for studies presenting data on the prevalence of heterozygous FH among different ethnicities in non-founder populations. Studies with more than 100 individuals, relevant data on prevalence, ethnicity, and using the Dutch Lipid Clinical Network Criteria, Simon Broome, Making Early Diagnosis Prevents Early Death, genetic screening, or comparable diagnostic criteria were considered eligible for inclusion.

Results: Eleven general population studies and two patient studies were included in a systematic review and 11 general population studies in a random-effects meta-analysis. The overall pooled FH prevalence was 0.33% or 1:303 in 1,169,879 individuals (95% confidence interval: 0.26–0:40%; 1:385–1:250). Included studies presented data on six ethnicities: black, Latino, white, Asian, brown, and mixed/other. Pooled prevalence was estimated for each group. The highest prevalence observed was 0.52% or 1:192 among blacks (0.34–0.69%; 1:294–1:145) and 0.48% or 1:208 among browns (0.31–0.74%; 1:323–1:135) while the lowest pooled prevalence was 0.25% or 1:400 among Asians (0.15–0.35; 1:500–1:286). The prevalence was 0.37% or 1:270 among Latino (0.24–0.69%; 1:417–1:145), 0.31% or 1:323 among white (0.24–0.41%; 1:417–1:244), and 0.32% or 1:313 among mixed/other individuals (0.13–0.52%; 1:769–1:192).

Conclusion: The estimated FH prevalence displays a variation across ethnicity, ranging from 0.25% (1:400) to 0.52% (1:192), with the highest prevalence seen among the black and brown and the lowest among the Asian individuals. The differences observed suggest that targeted screening among subpopulations may increase the identification of cases and thus the opportunity for prevention.

LDL-Cholesterol versus Glucose in Microvascular and Macrovascular Disease

BACKGROUND

The causal relationships between increased concentrations of low density lipoprotein (LDL)-cholesterol and glucose and risk of ischemic heart disease are well established. The causal contributions of LDL-cholesterol and glucose to risk of peripheral micro- and macrovascular diseases are less studied, especially in prediabetic stages and in a general population setting.

CONTENT

This review summarizes the current evidence for a causal contribution of LDL-cholesterol and glucose to risk of a spectrum of peripheral micro- and macrovascular diseases and reviews possible underlying disease mechanisms, including differences between vascular compartments, and finally discusses the clinical implications of these findings, including strategies for prevention and treatment.

SUMMARY

Combined lines of evidence suggest that LDL-cholesterol has a causal effect on risk of peripheral arterial disease and chronic kidney disease, both of which represent manifestations of macrovascular disease due to atherosclerosis and accumulation of LDL particles in the arterial wall. In contrast, there is limited evidence for a causal effect on risk of microvascular disease. Glucose has a causal effect on risk of both micro- and macrovascular disease. However, most evidence is derived from studies of individuals with diabetes. Further studies in normoglycemic and prediabetic individuals are warranted. Overall, LDL-cholesterol-lowering reduces risk of macrovascular disease, while evidence for a reduction in risk of microvascular disease is inconsistent. Glucose-lowering has a beneficial effect on risk of microvascular diseases and on risk of chronic kidney disease and estimated glomerular filtration rate (eGFR) in some studies, while results on risk of peripheral arterial disease are conflicting.

Impact of high glucose levels and glucose lowering on risk of ischaemic stroke: a Mendelian randomisation study and meta-analysis

AIMS/HYPOTHESIS

It is unclear whether glucose per se has a causal impact on risk of stroke and whether glucose-lowering drugs reduce this risk. This is important for the choice of treatment for individuals at risk. We tested the hypotheses that high plasma glucose has a causal impact on increased risk of ischaemic stroke, and that glucose-lowering drugs reduce this risk.

METHODS

Using a Mendelian randomisation design, we examined 118,838 individuals from two Copenhagen cohorts, the Copenhagen General Population Study and the Copenhagen City Heart Study, and 440,328 individuals from the MEGASTROKE study. Effects of eight glucose-lowering drugs on risk of stroke were summarised by meta-analyses.

RESULTS

In genetic, causal analyses, a 1 mmol/l higher plasma glucose had a risk ratio of 1.48 (95% CI 1.04, 2.11) for ischaemic stroke in the Copenhagen studies. The corresponding risk ratio from the MEGASTROKE study combined with the Copenhagen studies was 1.74 (1.31, 2.18). In meta-analyses of glucose-lowering drugs, the risk ratio for stroke was 0.85 (0.77, 0.94) for glucagon-like peptide-1 receptor agonists and 0.82 (0.69, 0.98) for thiazolidinediones, while sulfonylureas, dipeptidyl peptidase-4 inhibitors, sodium-glucose cotransporter 2 inhibitors, α-glucosidase inhibitors, meglitinides and metformin individually lacked statistical evidence of an effect on stroke risk.

CONCLUSIONS/INTERPRETATION

Genetically high plasma glucose has a causal impact on increased risk of ischaemic stroke. Treatment with glucose-lowering glucagon-like peptide-1 receptor agonists and thiazolidinediones reduces this risk. These results may guide clinicians in the treatment of individuals at high risk of ischaemic stroke.

Impact of Glucose Level on Micro- and Macrovascular Disease in the General Population: A Mendelian Randomization Study

OBJECTIVE

To evaluate whether high glucose levels in the normoglycemic range and higher have a causal genetic effect on risk of retinopathy, neuropathy, nephropathy, chronic kidney disease (CKD), peripheral arterial disease (PAD), and myocardial infarction (MI; positive control) in the general population.

RESEARCH DESIGN AND METHODS

This study applied observational and one-sample Mendelian randomization (MR) analyses to individual-level data from 117,193 Danish individuals, and validation by two-sample MR analyses on summary-level data from 133,010 individuals from the Meta-Analyses of Glucose and Insulin-Related Traits Consortium (MAGIC), 117,165 from the CKDGen Consortium, and 452,264 from the UK Biobank.

RESULTS

Observationally, glucose levels in the normoglycemic range and higher were associated with high risks of retinopathy, neuropathy, diabetic nephropathy, PAD, and MI (all P for trend <0.001). In genetic causal analyses, the risk ratio for a 1 mmol/L higher glucose level was 2.01 (95% CI 1.18-3.41) for retinopathy, 2.15 (1.38-3.35) for neuropathy, 1.58 (1.04-2.40) for diabetic nephropathy, 0.97 (0.84-1.12) for estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m², 1.19 (0.90-1.58) for PAD, and 1.49 (1.02-2.17) for MI. Summary-level data from the MAGIC, the CKDGen Consortium, and the UK Biobank gave a genetic risk ratio of 4.55 (95% CI 2.26-9.15) for retinopathy, 1.48 (0.83-2.66) for peripheral neuropathy, 0.98 (0.94-1.01) for eGFR <60 mL/min/1.73 m², and 1.23 (0.57-2.67) for PAD per 1 mmol/L higher glucose level.

CONCLUSIONS

Glucose levels in the normoglycemic range and higher were prospectively associated with a high risk of retinopathy, neuropathy, diabetic nephropathy, eGFR <60 mL/min/1.73 m², PAD, and MI. These associations were confirmed in genetic causal analyses for retinopathy, neuropathy, diabetic nephropathy, and MI, but they could not be confirmed for PAD and seemed to be refuted for eGFR <60 mL/min/1.73 m².

Familial Hypercholesterolemia and Risk of Peripheral Arterial Disease and Chronic Kidney Disease

CONTEXT

Individuals with familial hypercholesterolemia (FH) have a high risk of coronary artery disease, but their risk of peripheral arterial disease (PAD) and chronic kidney disease (CKD) is unknown.

OBJECTIVE

In individuals with clinical FH, we tested the hypotheses (1) that the risks of PAD and CKD are elevated and (2) that low ankle-brachial index (ABI) and estimated glomerular filtration rate (eGFR) are associated with a high risk of myocardial infarction.

DESIGN AND SETTING

Prospective cohort study of the general population.

PARTICIPANTS

A total of 106,172 individuals, of whom 7109 were diagnosed with FH.

MAIN OUTCOME MEASURES

PAD, CKD, and myocardial infarction.

RESULTS

Compared with individuals with unlikely FH, multivariable adjusted ORs (95% CIs) of PAD were 1.84 (1.70 to 2.00) in those with possible FH and 1.36 (1.00 to 1.84) in individuals with probable/definite FH. For CKD, the corresponding ORs (95% CIs) were 1.92 (1.78 to 2.07) and 2.42 (1.86 to 3.26). Compared with individuals with unlikely FH and ABI >0.9, the multivariable adjusted hazard ratio (95% CI) of myocardial infarction was 4.60 (2.36 to 8.97) in those with possible/probable/definite FH and ABI ≤0.9. Compared with individuals with unlikely FH and eGFR ≥60 mL/min/1.73 m2, the corresponding value was 2.19 (1.71 to 2.82) in those with possible/probable/definite FH and eGFR <60 mL/min/1.73 m2.

CONCLUSIONS

Individuals with clinical FH have increased risks of PAD and CKD, and low ABI and eGFR are associated with high risk of myocardial infarction. Consequently, individuals with FH should be screened for PAD and CKD, and ABI and eGFR may be used as prognostic tools in the management and treatment of FH to identify those at very high risk of myocardial infarction.

Bias due to lack of patient blinding in clinical trials. A systematic review of trials randomizing patients to blind and nonblind sub-studies

BACKGROUND

Blinding patients in clinical trials is a key methodological procedure, but the expected degree of bias due to nonblinded patients on estimated treatment effects is unknown.

METHODS

Systematic review of randomized clinical trials with one sub-study (i.e. experimental vs control) involving blinded patients and another, otherwise identical, sub-study involving nonblinded patients. Within each trial, we compared the difference in effect sizes (i.e. standardized mean differences) between the sub-studies. A difference <0 indicates that nonblinded patients generated a more optimistic effect estimate. We pooled the differences with random-effects inverse variance meta-analysis, and explored reasons for heterogeneity.

RESULTS

Our main analysis included 12 trials (3869 patients). The average difference in effect size for patient-reported outcomes was -0.56 (95% confidence interval -0.71 to -0.41), (I(2)=60%, P=0.004), i.e. nonblinded patients exaggerated the effect size by an average of 0.56 standard deviation, but with considerable variation. Two of the 12 trials also used observer-reported outcomes, showing no indication of exaggerated effects due lack of patient blinding. There was a larger effect size difference in 10 acupuncture trials [-0.63 (-0.77 to -0.49)], than in the two non-acupuncture trials [-0.17 (-0.41 to 0.07)]. Lack of patient blinding also increased attrition and use of co-interventions: ratio of control group attrition risk 1.79 (1.18 to 2.70), and ratio of control group co-intervention risk 1.55 (0.99 to 2.43).

CONCLUSIONS

This study provides empirical evidence of pronounced bias due to lack of patient blinding in complementary/alternative randomized clinical trials with patient-reported outcomes.

Observer bias in randomized clinical trials with time-to-event outcomes: systematic review of trials with both blinded and non-blinded outcome assessors

BACKGROUND

We wanted to evaluate the impact of nonblinded outcome assessors on estimated treatment effects in time-to-event trials.

METHODS

Systematic review of randomized clinical trials with both blinded and nonblinded assessors of the same time-to-event outcome. Two authors agreed on inclusion of trials and outcomes. We compared hazard ratios based on nonblinded and blinded assessments. A ratio of hazard ratios (RHR)<1 indicated that nonblinded assessors generated more optimistic effect estimates. We pooled RHRs with inverse variance random-effects meta-analysis.

RESULTS

We included 18 trials. Eleven trials (1969 patients) with subjective outcomes provided hazard ratios, RHR 0.88 (0.69 to 1.12), (I2=44%, P=0.06), but unconditional pooling was problematic because of qualitative heterogeneity. Four atypical cytomegalovirus retinitis trials compared experimental oral administration with control intravenous administration of the same drug, resulting in bias favouring the control intervention, RHR 1.33 (0.98 to 1.82). Seven trials of cytomegalovirus retinitis, tibial fracture and multiple sclerosis compared experimental interventions with standard control interventions, e.g. placebo, no-treatment or active control, resulting in bias favouring the experimental intervention, RHR 0.73 (0.57 to 0.93), indicating an average exaggeration of nonblinded hazard ratios by 27% (7% to 43%).

CONCLUSIONS

Lack of blinded outcome assessors in randomized trials with subjective time-to-event outcomes causes high risk of observer bias. Nonblinded outcome assessors typically favour the experimental intervention, exaggerating the hazard ratio by an average of approximately 27%; but in special situations, nonblinded outcome assessors favour control interventions, inducing a comparable degree of observer bias in the reversed direction.

Faecal microbiota transplantation and bacteriotherapy for recurrent Clostridium difficile infection: a retrospective evaluation of 31 patients

BACKGROUND

Recurrent Clostridium difficile infection (CDI) is a significant problem due to its increased incidence and severity. Failure rates for standard antibiotic therapies are high. In our hospital, faecal microbiota transplantation (FMT), or instillation of a culture mixture of known enteric bacteria in saline as rectal bacteriotherapy (RBT), has long been used as 'rescue therapy' in patients with recurrent disease, in whom repeated courses of standard antibiotic treatment have failed. We wanted to evaluate the effectiveness of FMT and RBT for recurrent CDI.

METHODS

The records of 31 patients treated with either FMT or RBT for recurrent CDI were reviewed retrospectively. FMT was based on faecal donation by a close relative and RBT on a defined saline mixture of 10 individually cultured enteric bacterial strains originally isolated from healthy persons. Both types of instillation were carried out through a rectal catheter. FMT (500 ml) was given as 1 installation. RBT (200 ml) was given as 2 or 3 installations with an interval of 2 days between courses. Treatment success was defined as a sustained loss of symptoms and discontinuation of diarrhoea within 3 days.

RESULTS

Of 31 patients, 23 (74%) responded successfully to the treatment: 16 of 23 (70%) receiving FMT and 7 of 8 (88%) receiving RBT.

CONCLUSION

We found FMT to be effective in patients with recurrent CDI. RBT based on a predefined bacterial suspension was as effective as or better than FMT based on faecal donation; however, multiple installations may be needed.

Observer bias in randomized clinical trials with measurement scale outcomes: a systematic review of trials with both blinded and nonblinded assessors

BACKGROUND

Clinical trials are commonly done without blinded outcome assessors despite the risk of bias. We wanted to evaluate the effect of nonblinded outcome assessment on estimated effects in randomized clinical trials with outcomes that involved subjective measurement scales.

METHODS

We conducted a systematic review of randomized clinical trials with both blinded and nonblinded assessment of the same measurement scale outcome. We searched PubMed, EMBASE, PsycINFO, CINAHL, Cochrane Central Register of Controlled Trials, HighWire Press and Google Scholar for relevant studies. Two investigators agreed on the inclusion of trials and the outcome scale. For each trial, we calculated the difference in effect size (i.e., standardized mean difference between nonblinded and blinded assessments). A difference in effect size of less than 0 suggested that nonblinded assessors generated more optimistic estimates of effect. We pooled the differences in effect size using inverse variance random-effects meta-analysis and used metaregression to identify potential reasons for variation.

RESULTS

We included 24 trials in our review. The main meta-analysis included 16 trials (involving 2854 patients) with subjective outcomes. The estimated treatment effect was more beneficial when based on nonblinded assessors (pooled difference in effect size -0.23 [95% confidence interval (CI) -0.40 to -0.06]). In relative terms, nonblinded assessors exaggerated the pooled effect size by 68% (95% CI 14% to 230%). Heterogeneity was moderate (I(2) = 46%, p = 0.02) and unexplained by metaregression.

INTERPRETATION

We provide empirical evidence for observer bias in randomized clinical trials with subjective measurement scale outcomes. A failure to blind assessors of outcomes in such trials results in a high risk of substantial bias.

Observer bias in randomised clinical trials with binary outcomes: systematic review of trials with both blinded and non-blinded outcome assessors

OBJECTIVE

To evaluate the impact of non-blinded outcome assessment on estimated treatment effects in randomised clinical trials with binary outcomes.

DESIGN

Systematic review of trials with both blinded and non-blinded assessment of the same binary outcome. For each trial we calculated the ratio of the odds ratios--the odds ratio from non-blinded assessments relative to the corresponding odds ratio from blinded assessments. A ratio of odds ratios <1 indicated that non-blinded assessors generated more optimistic effect estimates than blinded assessors. We pooled the individual ratios of odds ratios with inverse variance random effects meta-analysis and explored reasons for variation in ratios of odds ratios with meta-regression. We also analysed rates of agreement between blinded and non-blinded assessors and calculated the number of patients needed to be reclassified to neutralise any bias.

DATA SOURCES

PubMed, Embase, PsycINFO, CINAHL, Cochrane Central Register of Controlled Trials, HighWire Press, and Google Scholar.

ELIGIBILITY CRITERIA FOR SELECTING STUDIES

Randomised clinical trials with blinded and non-blinded assessment of the same binary outcome.

RESULTS

We included 21 trials in the main analysis (with 4391 patients); eight trials provided individual patient data. Outcomes in most trials were subjective--for example, qualitative assessment of the patient's function. The ratio of the odds ratios ranged from 0.02 to 14.4. The pooled ratio of odds ratios was 0.64 (95% confidence interval 0.43 to 0.96), indicating an average exaggeration of the non-blinded odds ratio by 36%. We found no significant association between low ratios of odds ratios and scores for outcome subjectivity (P=0.27); non-blinded assessor's overall involvement in the trial (P=0.60); or outcome vulnerability to non-blinded patients (P=0.52). Blinded and non-blinded assessors agreed in a median of 78% of assessments (interquartile range 64-90%) in the 12 trials with available data. The exaggeration of treatment effects associated with non-blinded assessors was induced by the misclassification of a median of 3% of the assessed patients per trial (1-7%).

CONCLUSIONS

On average, non-blinded assessors of subjective binary outcomes generated substantially biased effect estimates in randomised clinical trials, exaggerating odds ratios by 36%. This bias was compatible with a high rate of agreement between blinded and non-blinded outcome assessors and driven by the misclassification of few patients.