True Endpoint Research Articles

Differences in surrogate threshold effect estimates between original and simplified correlation-based validation approaches.

Surrogate endpoint validation has been well established by the meta-analytical correlation-based approach as outlined in the seminal work of Buyse et al. (Biostatistics, 2000). Surrogacy can be assumed if strong associations on individual and study levels can be demonstrated. Alternatively, if an effect on a true endpoint is to be predicted from a surrogate endpoint in a new study, the surrogate threshold effect (STE, Burzykowski and Buyse, Pharmaceutical Statistics, 2006) can be used. In practice, as individual patient data (IPD) are hard to obtain, some authors use only aggregate data and perform simplified regression analyses. We are interested in to what extent such simplified analyses are biased compared with the ones from a full model with IPD. To this end, we conduct a simulation study with IPD and compute STEs from full and simplified analyses for varying data situations in terms of number of studies, correlations, variances and so on. In the scenarios considered, we show that, for normally distributed patient data, STEs derived from ordinary (weighted) linear regression generally underestimate STEs derived from the original model, whereas meta-regression often results in overestimation. Therefore, if individual data cannot be obtained, STEs from meta-regression may be used as conservative alternatives, but ordinary (weighted) linear regression should not be used for surrogate endpoint validation.

Assumption free modeling and monitoring of batch processes

Modeling strategies currently in use for the monitoring of batch processes where multivariate data are available have some limitations, particularly for batches where the true starting or end point are not the same on an absolute time scale, or the batch progression varies among batches. In this paper, a method capturing these differences and allowing modeling and monitoring of batches in relative time is proposed. Using scores from principal component analysis (PCA) models as a feature space the new methodology is better able to handle the challenges usually experienced in batch analysis. The feasibility of the relative time approach is demonstrated using data from a chemical synthesis and a pharmaceutical drying process.

Causal Measures of the Treatment Effect Captured by Candidate Surrogate Endpoints

In randomized clinical trials, true clinical endpoints that are rare or difficult to measure can be costly and time consuming. Therefore, there is an increasing need to substitute true clinical endpoints with surrogate endpoints. However, candidate surrogate endpoints require appropriate validation. To evaluate the extent of a treatment effect (TE) captured by candidate surrogate endpoints, various surrogacy measures have been proposed by biostatisticians and medical professionals. Although many researchers stated that it is desirable that surrogacy measures should take values between zero to one, those often fall outside a range [0, 1] without suitable assumptions. To overcome this problem, we propose two types of surrogacy measures based on the causal association (CA) paradigm and the causal effect (CE) paradigm. These operate by decomposing the TE into those parts that are and are not captured by candidate surrogate endpoints. The surrogacy measures based on the CA paradigm mainly consider how much of the TE of the treatment on the true clinical endpoint can be predicted through the TE on the candidate surrogate endpoints, while the surrogacy measures based on the CE paradigm are concerned with how much of the TE on the true clinical endpoint is a result of the candidate surrogate endpoints. In addition, we demonstrate some properties of the proposed surrogacy measures, and show that they always fall inside the range [0, 1]. Furthermore, they can be considered as improved and extended versions of existing surrogacy measures. Based on simulation experiments and applications of the proposed surrogacy measures to a case study of the Olmesartan Reducing Incidence of End-stage Renal Disease in Diabetic Nephropathy Trial, we show that the proposed surrogacy measures solve the problems encountered by the existing surrogacy measures. This paper presents new quantitative surrogacy measures that reliably evaluate the proportion of the TE captured by candidate surrogate endpoints. Supplementary materials accompanying this paper appear online.

Assessing treatment effects with surrogate survival outcomes using an internal validation subsample.

In studies with surrogate outcomes available for all subjects and true outcomes available for only a subsample, survival analysis methods are needed that incorporate both endpoints in order to assess treatment effects. We develop a semiparametric estimated likelihood method for the proportional hazards model with discrete time data and a binary covariate of interest. Our proposed method allows for real-time validation of surrogate outcomes and flexible censoring mechanisms. Our proposed estimator is consistent and asymptotically normal. Through numerical studies, we showed that our proposed method for estimating a covariate effect is unbiased compared to the naïve estimator that uses only surrogate endpoints and is more efficient with moderate missingness compared to the complete-case estimator that uses only true endpoints. We further demonstrated the advantages of our proposed method in comparison with existing approaches when there is real-time validation. We also illustrated the use of our proposed method by estimating the effect of gender on time to detection of Alzheimer's disease using data from the Alzheimer's Disease Neuroimaging Initiative. The proposed method is able to account for the uncertainty of surrogate outcomes using a validation subsample of true outcomes in estimating a binary covariate effect. The proposed estimator can outperform standard semiparametric survival analysis methods and can therefore save on costs of a trial or improve power in detecting treatment effects.

The perils of surrogate endpoints.

Our goals in medicine are (i) to improve the quality of patients' lives, (ii) help them to live longer, and (iii) to do so at a reasonable cost. These are our true endpoints: health status, survival, and cost. It is thus entirely consistent with this point of view that these are the fundamental concepts that can be united in a formal cost–utility analysis.1 These endpoints remain the best measures of efficacy in clinical trials comparing a new therapy to placebo or to an active control. All other measures may then be seen as surrogate endpoints or surrogates. Thus, even serious events such as myocardial infarction and stroke may be considered surrogates, as their effect is to adversely affect the critical endpoints of health status, survival, and cost. However, in common usage a surrogate is a relatively easy to measure endpoint, available over a relatively short timeframe that is used in place of the true endpoints. Generally, surrogates are not events, but rather measurements (physiological, laboratory, or test results, e.g. biomarkers) that predict events. Thus, surrogates are most commonly measures that we can record, often with much shorter timescales than is necessary for events.2 Surrogates are usually continuous variables, often but not necessarily with reasonable approximations of a normal distribution. Continuous variables, especially if approximating a normal distribution, will allow for much smaller sample sizes than dichotomous variables as well as shorter periods of follow-up and lower costs. Thus, compared with clinical outcome trials, studies with surrogate endpoints can be conducted rapidly and with much less resource use and expense …

Ribosome profiling to investigate in vivo molecular mechanisms of immune function (TECH2P.906)

Abstract Genomic analyses of gene expression of immune cells have so far focused on their transcriptional program by quantifying mRNA abundance, measured by microarray or more recently by RNA-seq. The mRNA abundance is often presumed as a proxy of the gene expression, but the correlation between mRNA abundance and protein level, which is the true end point of gene expression, frequently shows lack of sufficient correlation due to the apparent regulation of gene expression at the layer of translation. Here, we present the applications of an emerging technique, Ribosome profiling (Ribo-seq), on primary immune cells derived from various genetic mutant mice developing immune diseases. This directly captured genome-wide maps of protein synthesis (translatome) by allowing quantification of ribosome footprint of each mRNA with high resolution and depth. Side-by-side comparison of transcriptome and translatome in the disease models revealed the signaling pathways that are predominantly regulated in translation levels are prevalent, highlighting the critical roles of translational control in immune function.

Statistical evaluation of surrogate endpoints with examples from cancer clinical trials.

A surrogate endpoint is intended to replace a clinical endpoint for the evaluation of new treatments when it can be measured more cheaply, more conveniently, more frequently, or earlier than that clinical endpoint. A surrogate endpoint is expected to predict clinical benefit, harm, or lack of these. Besides the biological plausibility of a surrogate, a quantitative assessment of the strength of evidence for surrogacy requires the demonstration of the prognostic value of the surrogate for the clinical outcome, and evidence that treatment effects on the surrogate reliably predict treatment effects on the clinical outcome. We focus on these two conditions, and outline the statistical approaches that have been proposed to assess the extent to which these conditions are fulfilled. When data are available from a single trial, one can assess the "individual level association" between the surrogate and the true endpoint. When data are available from several trials, one can additionally assess the "trial level association" between the treatment effect on the surrogate and the treatment effect on the true endpoint. In the latter case, the "surrogate threshold effect" can be estimated as the minimum effect on the surrogate endpoint that predicts a statistically significant effect on the clinical endpoint. All these concepts are discussed in the context of randomized clinical trials in oncology, and illustrated with two meta-analyses in gastric cancer.

Inference for Surrogate Endpoint Validation in the Binary Case

Surrogate endpoint validation for a binary surrogate endpoint and a binary true endpoint is investigated using the criteria of proportion explained (PE) and the relative effect (RE). The concepts of generalized confidence intervals and fiducial intervals are used for computing confidence intervals for PE and RE. The numerical results indicate that the proposed confidence intervals are satisfactory in terms of coverage probability, whereas the intervals based on Fieller’s theorem and the delta method fall short in this regard. Our methodology can also be applied to interval estimation problems in a causal inference-based approach to surrogate endpoint validation.

Surrogacy assessment using principal stratification with multivariate normal and Gaussian copula models.

The validation of intermediate markers as surrogate markers (S) for the true outcome of interest (T) in clinical trials offers the possibility for trials to be run more quickly and cheaply by using the surrogate endpoint in place of the true endpoint. Working within a principal stratification framework, we propose causal quantities to evaluate surrogacy using a Gaussian copula model for an ordinal surrogate and time-to-event final outcome. The methods are applied to data from four colorectal cancer clinical trials, where S is tumor response and T is overall survival. For the Gaussian copula model, a Bayesian estimation strategy is used and, as some parameters are not identifiable from the data, we explore the use of informative priors that are consistent with reasonable assumptions in the surrogate marker setting to aid in estimation. While there is some bias in the estimation of the surrogacy quantities of interest, the estimation procedure does reasonably well at distinguishing between poor and good surrogate markers. Some of the parameters of the proposed model are not identifiable from the data, and therefore, assumptions must be made in order to aid in their estimation. The proposed quantities can be used in combination to provide evidence about the validity of S as a surrogate marker for T.

Evaluating surrogate endpoints, prognostic markers, and predictive markers: Some simple themes.

A surrogate endpoint is an endpoint observed earlier than the true endpoint (a health outcome) that is used to draw conclusions about the effect of treatment on the unobserved true endpoint. A prognostic marker is a marker for predicting the risk of an event given a control treatment; it informs treatment decisions when there is information on anticipated benefits and harms of a new treatment applied to persons at high risk. A predictive marker is a marker for predicting the effect of treatment on outcome in a subgroup of patients or study participants; it provides more rigorous information for treatment selection than a prognostic marker when it is based on estimated treatment effects in a randomized trial. We organized our discussion around a different theme for each topic. "Fundamentally an extrapolation" refers to the non-statistical considerations and assumptions needed when using surrogate endpoints to evaluate a new treatment. "Decision analysis to the rescue" refers to use the use of decision analysis to evaluate an additional prognostic marker because it is not possible to choose between purely statistical measures of marker performance. "The appeal of simplicity" refers to a straightforward and efficient use of a single randomized trial to evaluate overall treatment effect and treatment effect within subgroups using predictive markers. The simple themes provide a general guideline for evaluation of surrogate endpoints, prognostic markers, and predictive markers.

Evaluation of an imaging biomarker, Dixon quantitative chemical shift imaging, in Gaucher disease: lessons learned.

Gaucher disease (GD) is the first lysosomal storage disorder for which specific therapy became available. The infiltration of bone marrow by storage cells plays an important part in the pathophysiology of skeletal complications and can be quantified by measurements of bone marrow fat fraction (Ff). Ff measurements by Dixon quantitative chemical shift imaging (QCSI) are standard for the follow-up care of GD patients at the Academic Medical Center. Several criteria should be met in order for these measurements to qualify as an imaging biomarker. These include: 1) The presence of the imaging biomarker is closely coupled or linked to the presence of the target disease or condition; 2) The detection and/or quantitative measurement of the biomarker is accurate, reproducible, and feasible over time, and; 3) The changes measured over time in the imaging biomarker are closely coupled, or linked, to the success or failure of the therapeutic effect and the true end point for the medical therapy being evaluated. This review assesses the use of Ff measurements by QCSI as a biomarker for GD in light of these criteria. In addition potential pitfalls are discussed including: degenerative disc disease; vertebral collapse and infection; haematological malignancies; focal fatty deposits; age; menopause; phase and repositioning errors, and; fat surrounding the basivertebral vein.QCSI measurements of Ff can be used as an imaging biomarker for GD taking these pitfalls into account. It is one of the first biomarkers, in particular imaging biomarkers, for GD that has been systematically evaluated and could be a valuable tool in clinical trials comparing different treatments or dosing regimens.

Semiparametric transformation models for semicompeting survival data.

Semicompeting risk outcome data (e.g., time to disease progression and time to death) are commonly collected in clinical trials. However, analysis of these data is often hampered by a scarcity of available statistical tools. As such, we propose a novel semiparametric transformation model that improves the existing models in the following two ways. First, it estimates regression coefficients and association parameters simultaneously. Second, the measure of surrogacy, for example, the proportion of the treatment effect that is mediated by the surrogate and the ratio of the overall treatment effect on the true endpoint over that on the surrogate endpoint, can be directly obtained. We propose an estimation procedure for inference and show that the proposed estimator is consistent and asymptotically normal. Extensive simulations demonstrate the valid usage of our method. We apply the method to a multiple myeloma trial to study the impact of several biomarkers on patients' semicompeting outcomes--namely, time to progression and time to death.

Odds Ratio for 2 × 2 Tables: Mantel–Haenszel Estimator, Profile Likelihood, and Presence of Surrogate Responses

Use of surrogate outcome to improve the inference in biomedical problems is an area of growing interest. Here, we consider a setup where both the true and surrogate endpoints are binary and we observe all the surrogate endpoints along with a few true endpoints. In a two-treatment setup we study the surrogate-augmented Mantel–Haenszel estimator based on observations from different groups when the group effect is present. We compare the Mantel–Haenszel estimator with the one obtained by maximizing profile likelihood in a surrogate augmented setup. We observe that the performances of these estimators are very close.

Ribosome profiling: new views of translation, from single codons to genome scale

Genome-wide analyses of gene expression have so far focused on the abundance of mRNA species as measured either by microarray or, more recently, by RNA sequencing. However, neither approach provides information on protein synthesis, which is the true end point of gene expression. Ribosome profiling is an emerging technique that uses deep sequencing to monitor in vivo translation. Studies using ribosome profiling have already provided new insights into the identity and the amount of proteins that are produced by cells, as well as detailed views into the mechanism of protein synthesis itself.

End Points in Lung Cancer Clinical Trials: Are we Ready to Step Away from Overall Survival?

467 ISSN 1758-1966 10.2217/LMT.13.54 © 2013 Future Medicine Ltd Lung Cancer Manage. (2013) 2(6), 467–470 Overall survival (OS), defined as the time from randomization or registration to death from any cause, is the gold standard end point in clinical trials as it is a measure of direct clinical benefit to a patient. OS as an end point is unambiguous and can unequivocally assess the benefit of a new treatment relative to the current standard of care. While improving OS remains the ultimate goal of new cancer therapy, an intermediate end point, such as progression-free survival (PFS), a composite end point defined as time from randomization to disease progression or death, is commonly used to evaluate the treatment effect of new oncologic products studied in randomized controlled trials (RCTs) [1]. Two major reasons to explore alternate end points to OS are: first, the inability to effectively assess crossover effects and subsequent therapies; and second, the requirement of large patient numbers and extended follow-up, making it a ‘long’ end point to assess, thereby delaying the development, approval and access to potentially efficacious drugs. Korn et al. discussed the influence of effective subsequent therapies on OS comparisons in RCTs where either subsequent therapy is equally effective in both arms of the RCT or where control arm patients crossover to experimental therapy upon disease progression [2]. The authors concluded that while improvement in an intermediate end point, such as PFS or disease-free survival (DFS), could be considered as evidence of true clinical benefit in adjuvant settings (i.e., early-stage disease), use of OS as the primary outcome is appropriate in trials where there is no intermediate end point to assess clinical benefit. This brings up the discussion of an intermediate end point versus a surrogate end point. An intermediate end point as defined by Korn et al. is one that directly measures true clinical benefit, whereas a surrogate end point is one that has been validated as an adequate substitute for the true clinical end point [2]. A valid surrogate end point must, therefore, both correlate with and accurately predict the true outcome of interest, for example, the treatment effect observed on the surrogate end point should reliably and precisely predict the treatment effect on the true clinical end point [3,4]. Shi et al. elegantly outline the issues surrounding the validation of surrogate end points [4], and review meta-analytic approaches

Statistical issues in the validation of prognostic, predictive, and surrogate biomarkers

Biomarkers have many distinct purposes, and depending on their intended use, the validation process varies substantially. The goal of this article is to provide an introduction to the topic of biomarkers, and then to discuss three specific types of biomarkers, namely, prognostic, predictive, and surrogate. A principle challenge for biomarker validation from a statistical perspective is the issue of multiplicity. In general, the solution to this multiplicity challenge is well known to statisticians: pre-specification and replication. Critical requirements for prognostic marker validation include uniform treatment, complete follow-up, unbiased case selection, and complete ascertainment of the many possible confounders that exist in the context of an observational sample. In the case of predictive biomarker validation, observational data are clearly inadequate and randomized controlled trials are mandatory. Within the context of randomization, strategies for predictive marker validation can be grouped into two categories: retrospective versus prospective validation. The critical validation criteria for a surrogate endpoint is to ensure that if a trial uses a surrogate endpoint, the trial will result in the same inferences as if the trial had observed the true endpoint. The field of surrogate endpoint validation has now moved to the multi-trial or meta-analytic setting as the preferred method. Biomarkers are a highly active research area. For all biomarker developmental and validation studies, the importance of fundamental statistical concepts remains the following: pre-specification of hypotheses, randomization, and replication. Further statistical methodology research in this area is clearly needed as we move forward.

Sharp bounds on causal effects using a surrogate endpoint

This paper considers a problem of evaluating the causal effect of a treatment X on a true endpoint Y using a surrogate endpoint S, in the presence of unmeasured confounders between S and Y. Such confounders render the causal effect of X on Y unidentifiable from the causal effect of X on S and the joint probability of S and Y. To evaluate the causal effect of X on Y in such a situation, this paper derives closed-form formulas for the sharp bounds on the causal effect of X on Y based on both the causal effect of X on S and the joint probability of S and Y under various assumptions. In addition, we show that it is not always necessary to observe Y to test the null causal effect of X on Y under the monotonicity assumption between X and S. These bounds enable clinical practitioners and researchers to assess the causal effect of a treatment on a true endpoint using a surrogate endpoint with minimum computational effort.

Computer-aided technology for fabricating complete dentures: Systematic review of historical background, current status, and future perspectives

Computer-aided technology is an emerging method for fabricating complete dentures. Consolidated information about historical background, current status, and scope for the future is lacking. The purpose of this systematic review was to analyze the existing literature on computer-aided technology for fabricating complete dentures and provide the reader with a historical background, current status, and future perspectives on this emerging technology. An electronic search of the English language literature between the periods of January 1957 and June 2012 was performed by using PubMed/MEDLINE with the following specific search terms: CAD-CAM complete dentures, digital complete dentures, computer dentures, designed dentures, machined dentures, manufactured dentures, milled dentures, and rapid prototyping dentures. Additionally, the search terms were used on the Google search engine to identify current commercial manufacturers and their protocols. A total of 1584 English language titles were obtained from the electronic database, and the systematic application of exclusion criteria resulted in the identification of 8 articles pertaining to computer-aided technology for complete dentures. Since the first published report in 1994, multiple authors have described different theoretical models and protocols for fabricating complete dentures with computer-aided technology. Although no clinical trials or clinical reports were identified in the scientific literature, the Google search engine identified 2 commercial manufacturers in the United States currently fabricating complete dentures with computer-aided design and computer-aided manufacturing (CAD/CAM) technology for clinicians world-wide. These manufacturers have definitive protocols in place and offer exclusive dental materials, techniques, and laboratory support. Their protocols contrast with conventional paradigms for fabricating complete dentures and allow the fabrication of complete dentures in 2 clinical appointments. A body of scientific literature related to computer-aided technology for complete dentures is emerging. Significant advancements in this technology have now resulted in their commercial availability with shorter clinical protocols. However, prospective clinical trials with true clinical endpoints are necessary to validate this technology. This could affect dental education, patient care, research, and public health worldwide.

Arterial Stiffness as an Imaging Biomarker

See related article, pp 161–167 The measurement of arterial stiffness is increasingly popular among physicians and researchers mainly because its predictive value for cardiovascular (CV) events has been well demonstrated. The largest amount of evidence has been given for aortic stiffness, measured through carotid-femoral pulse wave velocity (cfPWV). This has been initially reported in the late 1990s or early 2000s.1 Currently, as many as 19 studies consistently showed the predictive value of aortic stiffness for fatal and nonfatal CV events in various populations having different levels of CV risk: general population, patients with hypertension, elderly subjects, patients with type 2 diabetes mellitus, and patients with end-stage renal disease. In a recent meta-analysis,2 17 longitudinal studies totalizing 15 877 subjects with a mean follow-up of 7.7 years showed, for 1 SD increase in PWV, a risk ratio of 1.47 (1.31–1.64) for total mortality, 1.47 (1.29–1.66) for CV mortality, and 1.42 (1.29–1.58) for all-cause mortality. To be considered as a novel risk marker, arterial stiffness should add predictive information to established, standard risk markers, particularly the Framingham Risk Score or the European SCORE. This has been demonstrated with cfPWV in ≥3 studies: in patients with hypertension,3 in elderly subjects from a general population,4 and in middle-aged subjects from a general population.5 Another important requirement is that the measurement of the novel risk marker changes predicted risk sufficiently to change recommended therapy. This is indeed the case because several studies showed that a substantial amount of patients at intermediate risk could be reclassified into a higher or a lower CV risk, when arterial stiffness was measured.4–6 Aortic stiffness, measured through cfPWV, can thus be considered as a novel imaging biomarker for predicting CV events, although its value as a true surrogate end point requires a large intervention …

Hormonal versus non-hormonal contraceptives in women with diabetes mellitus type 1 and 2

Adequate contraceptive advice is important in both women with diabetes mellitus type 1 and type 2 to reduce the risk of maternal and infant morbidity and mortality in unplanned pregnancies. A wide variety of contraceptives are available for these women. However, hormonal contraceptives might influence carbohydrate and lipid metabolism and increase micro- and macrovascular complications, so caution in selecting a contraceptive method is required. To investigate whether progestogen-only, combined estrogen and progestogen or non-hormonal contraceptives differ in terms of effectiveness in preventing pregnancy, in their side effects on carbohydrate and lipid metabolism, and in long-term complications such as micro- and macrovascular disease when used in women with diabetes mellitus. The search was performed in CENTRAL, MEDLINE, EMBASE, POPLINE, CINAHL, WorldCat, ECO, ArticleFirst, the Science Citation Index, the British Library Inside, and reference lists of relevant articles. The last search was performed in January 2013. In addition, experts in the field and pharmaceutical companies marketing contraceptives were contacted to identify published, unpublished or ongoing studies. Randomised and quasi-randomised controlled trials that studied women with diabetes mellitus comparing: 1. hormonal versus non-hormonal contraceptives; 2. progestogen-only versus estrogen and progestogen contraceptives; 3. contraceptives containing < 50 µg estrogen versus contraceptives containing ≥ 50 µg estrogen; and 4. contraceptives containing first-, second- and third-generation progestogens, drospirenone and cyproterone acetate. The principal outcomes were contraceptive effectiveness, diabetes control, lipid metabolism and micro- and macrovascular complications. Two investigators evaluated the titles and abstracts identified from the literature search. Quality assessment was performed independently with discrepancies resolved by discussion or consulting a third review author. Because the trials differed in studied contraceptives, participant characteristics and methodological quality, we could not combine the data in a meta-analysis. The trials were therefore examined on an individual basis and narrative summaries were provided. Four randomised controlled trials were included. No unintended pregnancies were reported during the study periods. Only one trial was of good methodological quality. It compared the influence of a levonorgestrel-releasing intrauterine device (IUD) versus a copper IUD on carbohydrate metabolism in women with type 1 diabetes mellitus. No significant difference was found between the two groups. The other three trials were of limited methodological quality. Two compared progestogen-only pills with different estrogen and progestogen combinations, and one also included the levonorgestrel-releasing IUD and copper IUD. The trials reported that blood glucose levels remained stable during treatment with most regimens. Only high-dose combined oral contraceptives and 30 µg ethinylestradiol + 75 µg gestodene were identified as slightly impairing glucose homeostasis. The three studies found conflicting results regarding lipid metabolism. Some combined oral contraceptives appeared to have a minor adverse effect while others appeared to slightly improve lipid metabolism. The copper IUD and progestogen-only oral contraceptives also slightly improved lipid metabolism and no influence was seen while using the levonorgestel-releasing IUD. Only one study reported on micro- and macrovascular complications. It observed no signs or symptoms of thromboembolic incidents or visual disturbances, however study duration was short. Only minor adverse effects were reported in two studies. The four included randomised controlled trials in this systematic review provided insufficient evidence to assess whether progestogen-only and combined contraceptives differ from non-hormonal contraceptives in diabetes control, lipid metabolism and complications. Three of the four studies were of limited methodological quality, sponsored by pharmaceutical companies and described surrogate outcomes. Ideally, an adequately reported, high-quality randomised controlled trial analysing both intermediate outcomes (that is glucose and lipid metabolism) and true clinical endpoints (micro- and macrovascular disease) in users of combined, progestogen-only and non-hormonal contraceptives should be conducted. However, due to the low incidence of micro- and macrovascular disease and accordingly the large sample size and long follow-up period needed to observe differences in risk, a randomised controlled trial might not be the ideal design.