Prognostic modelling with logistic regression analysis: a comparison of selection and estimation methods in small data sets

Effectiveness of triclosan-coated PDS Plus versus uncoated PDS II sutures for prevention of surgical site infection after abdominal wall closure: the randomised controlled PROUD trial

BACKGROUND

Postoperative surgical site infections are one of the most frequent complications after open abdominal surgery, and triclosan-coated sutures were developed to reduce their occurrence. The aim of the PROUD trial was to obtain reliable data for the effectiveness of triclosan-coated PDS Plus sutures for abdominal wall closure, compared with non-coated PDS II sutures, in the prevention of surgical site infections.

METHODS

This multicentre, randomised controlled group-sequential superiority trial was done in 24 German hospitals. Adult patients (aged ≥18 years) who underwent elective midline abdominal laparotomy for any reason were eligible for inclusion. Exclusion criteria were impaired mental state, language problems, and participation in another intervention trial that interfered with the intervention or outcome of this trial. A central web-based randomisation tool was used to randomly assign eligible participants by permuted block randomisation with a 1:1 allocation ratio and block size 4 before mass closure to either triclosan-coated sutures (PDS Plus) or uncoated sutures (PDS II) for abdominal fascia closure. The primary endpoint was the occurrence of superficial or deep surgical site infection according to the Centers for Disease Control and Prevention criteria within 30 days after the operation. Patients, surgeons, and the outcome assessors were masked to group assignment. Interim and final analyses were by modified intention to treat. This trial is registered with the German Clinical Trials Register, number DRKS00000390.

FINDINGS

Between April 7, 2010, and Oct 19, 2012, 1224 patients were randomly assigned to intervention groups (607 to PDS Plus, and 617 to PDS II), of whom 1185 (587 PDS Plus and 598 PDS II) were analysed by intention to treat. The study groups were well balanced in terms of patient and procedure characteristics. The occurrence of surgical site infections did not differ between the PDS Plus group (87 [14·8%] of 587) and the PDS II group (96 [16·1%] of 598; OR 0·91, 95% CI 0·66-1·25; p=0·64). Serious adverse events also did not differ between the groups-146 of 583 (25·0%) patients treated with PDS Plus had at least one serious adverse event, compared with 138 of 602 (22·9%) patients treated with PDS II; p=0·39).

INTERPRETATION

Triclosan-coated PDS Plus did not reduce the occurrence of surgical site infection after elective midline laparotomy. Innovative, multifactorial strategies need to be developed and assessed in future trials to reduce surgical site infections.

FUNDING

Johnson & Johnson Medical Limited.

Prediction of survival with alternative modeling techniques using pseudo values

Background

The use of alternative modeling techniques for predicting patient survival is complicated by the fact that some alternative techniques cannot readily deal with censoring, which is essential for analyzing survival data. In the current study, we aimed to demonstrate that pseudo values enable statistically appropriate analyses of survival outcomes when used in seven alternative modeling techniques.

Methods

In this case study, we analyzed survival of 1282 Dutch patients with newly diagnosed Head and Neck Squamous Cell Carcinoma (HNSCC) with conventional Kaplan-Meier and Cox regression analysis. We subsequently calculated pseudo values to reflect the individual survival patterns. We used these pseudo values to compare recursive partitioning (RPART), neural nets (NNET), logistic regression (LR) general linear models (GLM) and three variants of support vector machines (SVM) with respect to dichotomous 60-month survival, and continuous pseudo values at 60 months or estimated survival time. We used the area under the ROC curve (AUC) and the root of the mean squared error (RMSE) to compare the performance of these models using bootstrap validation.

Results

Of a total of 1282 patients, 986 patients died during a median follow-up of 66 months (60-month survival: 52% [95% CI: 50%−55%]). The LR model had the highest optimism corrected AUC (0.791) to predict 60-month survival, followed by the SVM model with a linear kernel (AUC 0.787). The GLM model had the smallest optimism corrected RMSE when continuous pseudo values were considered for 60-month survival or the estimated survival time followed by SVM models with a linear kernel. The estimated importance of predictors varied substantially by the specific aspect of survival studied and modeling technique used.

Conclusions

The use of pseudo values makes it readily possible to apply alternative modeling techniques to survival problems, to compare their performance and to search further for promising alternative modeling techniques to analyze survival time.

Towards better clinical prediction models: seven steps for development and an ABCD for validation

Clinical prediction models provide risk estimates for the presence of disease (diagnosis) or an event in the future course of disease (prognosis) for individual patients. Although publications that present and evaluate such models are becoming more frequent, the methodology is often suboptimal. We propose that seven steps should be considered in developing prediction models: (i) consideration of the research question and initial data inspection; (ii) coding of predictors; (iii) model specification; (iv) model estimation; (v) evaluation of model performance; (vi) internal validation; and (vii) model presentation. The validity of a prediction model is ideally assessed in fully independent data, where we propose four key measures to evaluate model performance: calibration-in-the-large, or the model intercept (A); calibration slope (B); discrimination, with a concordance statistic (C); and clinical usefulness, with decision-curve analysis (D). As an application, we develop and validate prediction models for 30-day mortality in patients with an acute myocardial infarction. This illustrates the usefulness of the proposed framework to strengthen the methodological rigour and quality for prediction models in cardiovascular research.

Blood glucose as a marker of venous thromboembolism in critically ill children

BACKGROUND

The ability to predict the development of venous thromboembolism is highly desirable.

OBJECTIVE

We aim to determine the association between hyperglycemia and venous thromboembolism in non-diabetic critically ill children.

PATIENTS/METHODS

We conducted a retrospective cohort study that included children in the pediatric intensive care unit on a vasopressor or mechanical ventilator and without history of diabetes mellitus or prior diagnosis of thrombosis. Based on maximum blood glucose > 150 mg dL(-1) while admitted to the unit, children were categorized as hyperglycemic or non-hyperglycemic. The primary outcome was development of venous thromboembolism while admitted to the unit. We determined the association between hyperglycemia and venous thromboembolism using logistic regression models adjusting for selected subject characteristics.

RESULTS

Of the 789 subjects analyzed, 34 subjects developed venous thromboembolism (incidence, 4.3%; 95% confidence interval, 3-6%). Venous thromboembolism was more likely to develop in hyperglycemic subjects compared with non-hyperglycemic subjects. A total of 31 subjects (6.2%; 95% confidence interval, 4.2-8.7%) developed venous thromboembolism after becoming hyperglycemic compared with three non-hyperglycemic subjects with venous thromboembolism (1%; 95% confidence interval, 0.2-3%). When adjusted for age, diagnosis, presence of central venous catheter, prophylactic antithrombotic use and severity of illness, the odds ratio of venous thromboembolism with hyperglycemia was 4.1 (95% confidence interval, 1.2-14.1). For every 10 mg dL(-1) increase in maximum blood glucose, the adjusted odds ratio of venous thromboembolism was 1.04 (95% confidence interval, 1.01-1.06).

CONCLUSION

Hyperglycemia is associated with venous thromboembolism in critically ill non-diabetic children. Maximum blood glucose is a potential predictor of venous thromboembolism in this population.

Time to relapse following treatment for methamphetamine use: a long-term perspective on patterns and predictors

INTRODUCTION

This paper describes methamphetamine (MA) use patterns, specifically the duration of continuing abstinence ("time to relapse") for periods averaging 5 years post-discharge from treatment for MA use, and the relationship with selected user and treatment characteristics.

METHODS

A sample of 350 treatment admissions from a large county substance use disorder (SUD) treatment system was randomly selected (within gender, race/ethnicity, treatment modality strata). Retrospective self-report data are from natural history interviews (NHI) conducted approximately 3 years after treatment and a follow-up of 2-3 years later. Relapse is defined as any use of MA with time as the number of months of continuous MA abstinence after treatment discharge until relapse. This outcome was constructed from a monthly MA use timeline using NHI data. A Cox model was used to examine time to relapse and predictors.

RESULTS

Sixty-one percent of the sample relapsed to MA use within 1 year after treatment discharge and 14% during years 2-5. Significant protective factors predicting longer time to relapse included having experienced serious MA-related psychiatric/behavioral problems (hazard ratio [HR]=0.75, p=0.027), longer duration of the index treatment episode (HR=0.93, p=0.001), and participating in self-help or other treatment during the post-treatment abstinence period (HR=0.29, p<0.001); risk factors for shorter time to relapse included having a parent with alcohol and/or drug use problems (HR=1.35, p=0.020) and involvement in MA sales (HR=1.48, p=0.002).

CONCLUSIONS

Results contribute a long-term perspective on patterns of MA use following treatment and support a need for early post-treatment and long-term continuing care and relapse-prevention services.

Profiling Plasma MicroRNA in Nasopharyngeal Carcinoma with Deep Sequencing

BACKGROUND

The goal of this study was to establish a plasma microRNA profile by use of next-generation sequencing that could aid in assessment of patient prognosis in nasopharyngeal carcinoma (NPC).

METHODS

Two panels of NPC patients and healthy controls (HCs) were recruited for this study. We used deep sequencing to screen plasma microRNAs. Differentially expressed microRNAs were verified by quantitative real-time PCR (qPCR). Kaplan–Meier survival analysis with the log-rank test was used to compare overall survival (OS) and progression-free survival (PFS) between groups.

RESULTS

Twenty-three plasma miRNAs with differential expression levels were selected for qPCR analysis on an independent set including 100 NPC patients and 55 HCs. NPC patients with low concentrations of miR-483–5p and miR-103 had better prognosis for 5-year OS than those with high concentrations (87.5% vs 55.8%, P &lt; 0.001; 80.9% vs 62.3%, P = 0.031). Those with low concentrations of miR-29a and let-7c had poorer prognosis (54.8% vs 82.8%, P = 0.002; 56.3% vs 84.6%, P = 0.001). A 3-signature miRNA integrated with clinical stage was further identified in an independent set. We calculated a prognostic index score and classified patients into low-, medium-, and high-risk groups. Five-year OS among the 3 groups was significantly different (90.9%, 66.7%, and 23.8%; P &lt; 0.001). By multivariate analysis, a high-risk score was the most significantly unfavorable prognostic factor independent of other clinical variables (P &lt; 0.001, hazard ratio = 15.1, 95% CI = 5.2–43.9).

CONCLUSIONS

Differentially expressed plasma miRNAs as identified by next-generation sequencing can be helpful for predicting survival in NPC patients.

Predictors to parental knowledge about childhood immunisation/EPI vaccines in two health districts in Cameroon prior to the introduction of 13-valent Pneumococcal Conjugate Vaccines (PCV-13)

INTRODUCTION

Pneumonia is vaccine-preventable, but the increasing death toll resulting from the disease in Sub-Saharan Africa is alarming. Several factors account for vaccine failing to reach every child, besides incomplete vaccine coverage. Most of these include the perceptions of parents/guardians and healthcare providers. Previous studies on the introduction of new vaccines have focused on experimental trials, coverage figures and vaccine efficacy in developed countries. Little is known on the factors which may hinder the implementation process despite the huge challenges this may encounter in developing countries. This study described the knowledge, attitude and practices (KAP) of parents/guardians on pneumonia and immunisations/EPI vaccines; identify predictive parental socio-economic/demographic characteristics that of good knowledge on pneumonia infections, routine EPI vaccines and the PCV-13. Finally, the study described health center personnel perceptions about immunisations.

METHODS

The WHO's immunisation coverage cluster survey design was used, involving parents/guardians (n = 205) of children aged 0-59 months and health centre personnel (n = 13) directly concerned with vaccination activities between July-September 2010 in two health districts in Yaounde, Cameroon. Descriptive statistics and multivariate logistic models were used to analyse the parental/guardian data while the health personnel data was only analysed descriptively using SPSS version 17.0.

RESULTS

Only 19% of the parents/guardians were aware of the availability of the PCV-13. Logistic modelling identified important associations between parental socio-economic/demographic factors and good knowledge on pneumonia disease burden and prevention.

CONCLUSION

According to parents/guardians a short and clear message on the dangers of pneumonia and the need for prevention provided to parents/guardians during sensitisation/out-reach campaigns and use of social network avenues would be primordial, if the PCV-13 is to reach every child.

Improved diabetes screening using an extended predictive feature search

BACKGROUND

Screening entire populations for diabetes is not cost-effective. Hence, an efficient screening process must select those people who are at high risk for diabetes. In this study, we investigated whether screening procedures could be improved using an extended predictive feature search.

MATERIALS AND METHODS

In order to develop our model and identify persons with diabetes (prevalence) we used data from years of the National Health and Nutrition Examination Survey (2005-2010), which has not been explored for this purpose before. We calculated all combinations of predictors in order to identify the optimal subset, and we used a linear logistic classification model to predict diabetes. V-fold cross-validation was used for the process of including variables and for validating the final models. This new model was compared with two established models.

RESULTS

In total, 5,398 participants were included in this study. Among these, 478 participants had unidentified diabetes. The established models had a receiver operating characteristics curve for the area under the curve (AUC) of 0.74 and 0.71 compared with an AUC of 0.78 for the new model, showing a significant difference (P<0.05). A proposed cutoff point for the established models yielded respective sensitivities/specificities of 63%/72% and 40%/72% compared with the new model, which had a sensitivity/specificity of 70%/72%.

CONCLUSIONS

Our data indicate that simple healthcare and economic information such as ratio of family income to poverty can add value in deciding who is at risk of unknown diabetes by using extended investigations of predictor combinations.

Risk prediction with machine learning and regression methods

This is a discussion of issues in risk prediction based on the following papers: "Probability estimation with machine learning methods for dichotomous and multicategory outcome: Theory" by Jochen Kruppa, Yufeng Liu, Gérard Biau, Michael Kohler, Inke R. König, James D. Malley, and Andreas Ziegler; and "Probability estimation with machine learning methods for dichotomous and multicategory outcome: Applications" by Jochen Kruppa, Yufeng Liu, Hans-Christian Diener, Theresa Holste, Christian Weimar, Inke R. König, and Andreas Ziegler.

Attitudes and views of the general public towards research participation

BACKGROUND

While the challenges of recruitment into clinical trials are well described, little is known about the public's perceptions towards research.

AIMS

We sought to describe the attitudes, beliefs and knowledge of the public towards research and research participation, focusing on clinical trials, contrast these attributes among individuals with different relationships with the healthcare system and to identify predictors of willingness to participate.

METHODS

We conducted a self-administered cross-sectional survey of patients and their significant others in two clinics and two intensive care unit waiting rooms and in three public venues.

RESULTS

We analysed responses from 417 respondents (102 and 105 in dialysis and oncology clinics, and 106 in intensive care unit (ICU) waiting rooms, 104 in public locations). While most (68.3%) respondents favoured the use of humans in clinical trials, 53% felt that trial participants always or almost always receive the best quality of care, only 30.4% had participated in clinical research. Approximately 70% felt that subjects are always advised of the risks and benefits of participation, and 30% expressed ambiguity regarding whether participants are informed of their involvement. Oncology and dialysis respondents were the most and least informed regarding research methods and ethics. The perceived risks and benefits associated with clinical circumstances influence research participation decisions and vary with healthcare experiences. We identified six predictors of willingness to participate.

CONCLUSION

Attitudes of the public towards research participation are beleaguered by misconceptions. Stakeholders in clinical research must educate the general public regarding research methods and ethics.

Uncomplicated oocyte donation pregnancies are associated with a higher incidence of human leukocyte antigen alloantibodies

BACKGROUND

Fetuses in pregnancies conceived after oocyte donation (OD) have a higher degree of antigeneic dissimilarity with the mother compared to semi-allogeneic fetuses after natural conception. We questioned whether this leads to higher level of HLA antibody formation in OD pregnancies.

METHOD

Uncomplicated pregnancies after OD were compared with pregnancies conceived either spontaneously or by IVF. We calculated the number of HLA- and epitope mismatches. Maternal sera were screened for HLA antibodies with ELISA; child HLA specific antibody production was determined using CDC and Luminex with single antigen beads for class I and II.

RESULTS

A significantly (p<0.0001) higher incidence of HLA antibody production was observed in women conceiving after OD (69%) compared to non-donor pregnancies (24-25%). The antibody formation was positively correlated with the number of fetomaternal antigen (Spearman's rho 0.95, p<0.0001) and epitope mismatches (Spearman's rho 0.91, p<0.0001). The number of HLA-DR mismatches between women and child was an independent risk factor for the production of HLA class I specific alloantibodies.

CONCLUSION

Women conceiving after OD have a higher risk of developing child-specific HLA antibodies; the higher the number of immunogenetic differences, the higher the chance these antibodies are formed. The high incidence of antibody production also strongly depends upon the number of HLA-DR mismatches. Despite the stronger antibody response, OD was associated with uncomplicated pregnancy in cases included in this study.

Factors affecting employment among informal caregivers assisting people with multiple sclerosis

The objective of this study was to identify characteristics of informal caregivers, caregiving, and the people with multiple sclerosis (MS) receiving assistance that are associated with reduced caregiver employment. Data were collected during telephone interviews with 530 MS caregivers, including 215 employed caregivers, with these survey data analyzed using logistic regression. Poorer cognitive ability by the care recipient to make decisions about daily tasks and more caregiving hours per week predicted reduced caregiver employment. Better physical health domains of caregiver quality of life were associated with significantly lower odds of reduced employment. Health professionals treating informal caregivers, as well as those treating people with MS, need to be aware of respite, support, and intervention programs available to MS caregivers and refer them to these programs, which could reduce the negative impact of caregiving on employment.

The need for mental health care among informal caregivers assisting people with multiple sclerosis

The objective of this study was to identify characteristics of informal caregivers and people with multiple sclerosis (MS) receiving assistance that are associated with the caregiver's perceived need for mental health care. Survey data were collected in interviews with 530 caregivers and analyzed using a logistic regression model. We found that older caregiver age significantly decreased the odds of caregivers' perceived need for mental health treatment. Better mental health domains of health-related quality of life among caregivers, as measured by the 8-item Short Form Health Status Survey (SF-8), also were associated with decreased odds of the need for mental health care. In contrast, the caregiver's feeling that providing assistance was emotionally draining or the belief that this assistance threatened the caregiver/care recipient relationship significantly increased the odds of caregivers' needing mental health treatment. Health professionals treating informal caregivers should be sensitive to the impact that providing assistance has on the emotions, relationships, and mental health needs of caregivers.

Meta-analysis and aggregation of multiple published prediction models

Published clinical prediction models are often ignored during the development of novel prediction models despite similarities in populations and intended usage. The plethora of prediction models that arise from this practice may still perform poorly when applied in other populations. Incorporating prior evidence might improve the accuracy of prediction models and make them potentially better generalizable. Unfortunately, aggregation of prediction models is not straightforward, and methods to combine differently specified models are currently lacking. We propose two approaches for aggregating previously published prediction models when a validation dataset is available: model averaging and stacked regressions. These approaches yield user-friendly stand-alone models that are adjusted for the new validation data. Both approaches rely on weighting to account for model performance and between-study heterogeneity but adopt a different rationale (averaging versus combination) to combine the models. We illustrate their implementation in a clinical example and compare them with established methods for prediction modeling in a series of simulation studies. Results from the clinical datasets and simulation studies demonstrate that aggregation yields prediction models with better discrimination and calibration in a vast majority of scenarios, and results in equivalent performance (compared to developing a novel model from scratch) when validation datasets are relatively large. In conclusion, model aggregation is a promising strategy when several prediction models are available from the literature and a validation dataset is at hand. The aggregation methods do not require existing models to have similar predictors and can be applied when relatively few data are at hand.

Risk stratification for cardiac complications in patients hospitalized for community-acquired pneumonia

OBJECTIVE

To derive and validate a clinical rule that stratifies the risk of cardiac complications in patients hospitalized for community-acquired pneumonia (CAP) and compare its performance to the pneumonia severity index (PSI) score.

PATIENTS AND METHODS

Two cohorts of patients hospitalized for CAP were selected for the study. We used regression techniques in the derivation cohort (1343 patients enrolled in the Pneumonia Patient Outcomes Research Team study between October 1991 and March 1994) to generate a prediction rule that we validated in the validation cohort (608 patients enrolled in the Dissemination of Guidelines for Length of Stay study between February 1998 and March 1999). Discrimination and reclassification analyses compared its performance against the PSI score.

RESULTS

A prediction model for cardiac complications in the derivation cohort included age, 3 preexisting conditions, 2 vital signs, and 7 common laboratory or radiographic parameters. Discrimination (C statistic, 0.81; 95% CI, 0.78-0.84) and calibration (Hosmer-Lemeshow goodness-of-fit test, χ(2)=13.0; P=.11) were good. We derived a point score system from this model that when applied to the validation cohort also had good discrimination (C statistic, 0.78; 95% CI, 0.74-0.83) and calibration (Hosmer-Lemeshow, χ(2)=9.0; P=.34). On the basis of this score, we defined 4 categories of incremental risk of cardiac complications. The incidence of cardiac complications across risk categories increased linearly (from lowest to highest) in both the derivation (3.0%, 17.8%, 35.2%, and 72.2%) and validation (5.0%, 8.2%, 28.3%, and 48.9%) cohorts (Cochran-Armitage linear trend test, P<.01). The new score outperformed the PSI score in predicting cardiac complications in the validation cohort (C statistic, 0.78 vs 0.74; P=.03; proportion of patients correctly reclassified by the new score, 44%).

CONCLUSION

We derived and validated a clinical rule that accurately stratifies the risk of cardiac complications in patients hospitalized for CAP.

Comparison between various fracture risk assessment tools

The suboptimal performance of bone mineral density as the sole predictor of fracture risk and treatment decision making has led to the development of risk prediction algorithms that estimate fracture probability using multiple risk factors for fracture, such as demographic and physical characteristics, personal and family history, other health conditions, and medication use. We review theoretical aspects for developing and validating risk assessment tools, and illustrate how these principles apply to the best studied fracture probability tools: the World Health Organization FRAX®, the Garvan Fracture Risk Calculator, and the QResearch Database's QFractureScores. Model development should follow a systematic and rigorous methodology around variable selection, model fit evaluation, performance evaluation, and internal and external validation. Consideration must always be given to how risk prediction tools are integrated into clinical practice guidelines to support better clinical decision making and improved patient outcomes. Accurate fracture risk assessment can guide clinicians and individuals in understanding the risk of having an osteoporosis-related fracture and inform their decision making to mitigate these risks.

Predicting the severity of acute bronchiolitis in infants: should we use a clinical score or a biomarker?

Krebs von den Lungen 6 antigen (KL-6) has been shown to be a useful biomarker of the severity of Respiratory syncytial virus bronchiolitis. To assess the correlation between the clinical severity of acute bronchiolitis, serum KL-6, and the causative viruses, 222 infants with acute bronchiolitis presenting at the Pediatric Emergency Department of Estaing University Hospital, Clermont-Ferrand, France, were prospectively enrolled from October 2011 to May 2012. Disease severity was assessed with a score calculated from oxygen saturation, respiratory rate, and respiratory effort. A nasopharyngeal aspirate was collected to screen for a panel of 20 respiratory viruses. Serum was assessed and compared with a control group of 38 bronchiolitis-free infants. No significant difference in KL-6 levels was found between the children with bronchiolitis (mean 231 IU/mL ± 106) and those without (230 IU/mL ± 102), or between children who were hospitalized or not, or between the types of virus. No correlation was found between serum KL-6 levels and the disease severity score. The absence of Human Rhinovirus was a predictive factor for hospitalization (OR 3.4 [1.4-7.9]; P = 0.006). Older age and a higher oxygen saturation were protective factors (OR 0.65[0.55-0.77]; P < 0.0001 and OR 0.67 [0.54-0.85] P < 0.001, respectively). These results suggest that in infants presenting with bronchiolitis for the first time, clinical outcome depends more on the adaptive capacities of the host than on epithelial dysfunction intensity. Many of the features of bronchiolitis are affected by underlying disease and by treatment.

Miscount incidents: a novel approach to exploring risk factors for unintentionally retained surgical items

BACKGROUND

An estimated 1,500 operations result in retained surgical items (RSIs) each year in the United States, resulting in substantial morbidity. The rarity of these events makes studying them difficult, but miscount incidents may provide a window into understanding risk factors for RSIs.

METHODS

A cohort study of all consecutive operative cases during a 12-month period was conducted at a large academic medical center to identify risk factors for surgical miscounts. A multidisciplinary electronic miscount reconciliation checklist (necessitating both surgeon and nurse input) was introduced into the internally developed electronic Perioperative Information Management System to build a predictive model for RSI cases.

RESULTS

Among 23,955 operations, 84 resulted in miscount incidents (0.35% [95% confidence interval: 0.28% to 0.43%]). Increased case duration was strongly associated with increased risk of a miscount in unadjusted analyses (p < .0001). In the nested case-control analysis, both the case duration and the number of providers present were independently associated with a more than doubling of the odds of a miscount, even after adjustment for one another, the elective/urgent/emergent status of a case, and personnel changes occurring during the case.

CONCLUSIONS

The finding that both the length of the case and the number of providers involved in the case were independent risk factors for miscount incidents may offer insight into risk-targeted strategies to prevent RSIs, such as postoperative imaging, bar-coded surgical items, and radiofrequency technology. Miscounts trigger use of the Incorrect Count Safety Checklist, which can be used to determine whether a count completed at the procedure's conclusion is consistent across disciplines (circulating nurses, scrub persons, surgeons).

Principal component modeling of isokinetic moment curves for discriminating between the injured and healthy knees of unilateral ACL deficient patients

Bilateral knee strength evaluations of unilateral anterior cruciate ligament (ACL) deficient patients using isokinetic dynamometry are commonly performed in rehabilitation settings. The most frequently-used outcome measure is the peak moment value attained by the knee extensor and flexor muscle groups. However, other strength curve features may also be of clinical interest and utility. The purpose of this investigation was to identify, using Principal Component Analysis (PCA), strength curve features that explain the majority of variation between the injured and uninjured knee, and to assess the capabilities of these features to detect the presence of injury. A mixed gender cohort of 43 unilateral ACL deficient patients performed 6 continuous concentric knee extension and flexion repetitions bilaterally at 60°s(-1) and 180°s(-1) within a 90° range of motion. Moment waveforms were analyzed using PCA, and binary logistic regression was used to develop a discriminatory decision rule. For all directions and speeds, a statistically significant overall reduction in strength was noted for the involved knee in comparison to the uninvolved knee. The discriminatory decision rule yielded a specificity and sensitivity of 60.5% and 60.5%, respectively, corresponding to an accuracy of ∼62%. As such, the curve features extracted using PCA enabled only limited clinical usefulness in discerning between the ACL deficient and contra lateral, healthy knee. Improvement in discrimination capabilities may perhaps be achieved by consideration of different testing speeds and contraction modes, as well as utilization of other data analysis techniques.

Differences between women and men with recent-onset axial spondyloarthritis: results from a prospective multicenter French cohort

OBJECTIVE

To clarify sex differences in early axial spondyloarthritis (SpA).

METHODS

In total, 475 patients included in the Devenir des Spondylarthropathies Indifférenciées Récentes (Outcome of Recent Undifferentiated Spondylarthropathies) cohort, a prospective multicenter French cohort of patients with early inflammatory back pain suggestive of SpA, and fulfilling the Assessment of SpondyloArthritis international Society (ASAS) classification criteria for axial SpA were studied. The clinical and imaging features were compared between sexes and according to the clinical or imaging arm of the ASAS criteria using univariate and multivariate analysis.

RESULTS

Comparisons between the 239 men and 236 women showed that women had higher disease activity when measured by the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) and Bath Ankylosing Spondylitis Patient Global Score and higher fatigue and functional scores despite having less radiographic sacroiliitis and magnetic resonance imaging (MRI) inflammation of sacroiliac joints and the spine than men. Disease activity measured by the C-reactive protein (CRP)-based Ankylosing Spondylitis Disease Activity Score was not different between men and women. In contrast to patients classified with the clinical arm, disease activity and functional scores did not differ between women and men with sacroiliitis on imaging scans, except for fatigue and the Ankylosing Spondylitis Quality of Life questionnaire. Women with sacroiliitis had more peripheral involvement and more family history, whereas HLA-B27 positivity, elevated CRP, and MRI inflammation of the spine were associated with male sex.

CONCLUSION

Women with early axial SpA according to the ASAS criteria had greater disease activity when measured by the BASDAI and worse functioning despite fewer radiologic abnormalities than men. The differences in disease expression may be confounding factors to establish the diagnosis of SpA and to assess disease activity in women, suggesting that the imaging arm is a pivotal measure in the ASAS criteria.

Hip and knee replacement patients prefer pen-and-paper questionnaires: Implications for future patient-reported outcome measure studies

Objectives

Electronic forms of data collection have gained interest in recent years. In orthopaedics, little is known about patient preference regarding pen-and-paper or electronic questionnaires. We aimed to determine whether patients undergoing total hip (THR) or total knee replacement (TKR) prefer pen-and-paper or electronic questionnaires and to identify variables that predict preference for electronic questionnaires.

Methods

We asked patients who participated in a multi-centre cohort study investigating improvement in health-related quality of life (HRQoL) after THR and TKR using pen-and-paper questionnaires, which mode of questionnaire they preferred. Patient age, gender, highest completed level of schooling, body mass index (BMI), comorbidities, indication for joint replacement and pre-operative HRQoL were compared between the groups preferring different modes of questionnaire. We then performed logistic regression analyses to investigate which variables independently predicted preference of electronic questionnaires.

Results

A total of 565 THR patients and 387 TKR patients completed the preference question. Of the THR patients, 81.8% (95% confidence interval (CI) 78.4 to 84.7) preferred pen-and-paper questionnaires to electronic questionnaires, as did 86.8% (95% CI 83.1 to 89.8) of TKR patients. Younger age, male gender, higher completed level of schooling and higher BMI independently predicted preference of electronic questionnaires in THR patients. Younger age and higher completed level of schooling independently predicted preference of electronic questionnaires in TKR patients.

Conclusions

The majority of THR and TKR patients prefer pen-and-paper questionnaires. Patients who preferred electronic questionnaires differed from patients who preferred pen-and-paper questionnaires. Restricting the mode of patient-reported outcome measures to electronic questionnaires might introduce selection bias.

Cite this article: Bone Joint Res 2013;2:238–44.

Predictive performance of comorbidity measures in administrative databases for diabetes cohorts

BACKGROUND

The performance of comorbidity measures for predicting mortality in chronic disease populations and using ICD-9 diagnosis codes in administrative health data has been investigated in several studies, but less is known about predictive performance with ICD-10 data and for other health outcomes. This study investigated predictive performance of five comorbidity measures for population-based diabetes cohorts in administrative data. The objectives were to evaluate performance for: (a) disease-specific and general health outcomes, (b) data based on the ICD-9 and ICD-10 diagnoses, and (c) different age groups.

METHODS

Performance was investigated for heart attack, stroke, amputation, renal disease, hospitalization, and death in all-age and age-specific cohorts. Hospital records, physician billing claims, and prescription drug records from one Canadian province were used to identify diabetes cohorts and measure comorbidity. The data were analysed using multiple logistic regression models and summarized using measures of discrimination, accuracy, and fit.

RESULTS

In Cohort 1 (n = 29,058), for which only ICD-9 diagnoses were recorded in administrative data, the Elixhauser index showed good or excellent prediction for amputation, renal disease, and death and performed better than the Charlson index. Number of diagnoses was a good predictor of hospitalization. Similar results were obtained for Cohort 2 (n = 41,925), in which both ICD-9 and ICD-10 diagnoses were recorded in administrative data, although predictive performance was sometimes higher. For age-specific models of mortality, the Elixhauser index resulted in the largest improvement in predictive performance in all but the youngest age group.

CONCLUSIONS

Cohort age and the health outcome under investigation, but not the diagnosis coding system, may influence the predictive performance of comorbidity measure for studies about diabetes populations using administrative health data.

Low primary and secondary HIV drug-resistance after 12 months of antiretroviral therapy in human immune-deficiency virus type 1 (HIV-1)-infected individuals from Kigali, Rwanda

Treatment outcomes of HIV patients receiving antiretroviral therapy (ART) in Rwanda are scarcely documented. HIV viral load (VL) and HIV drug-resistance (HIVDR) outcomes at month 12 were determined in a prospective cohort study of antiretroviral–naïve HIV patients initiating first-line therapy in Kigali. Treatment response was monitored clinically and by regular CD4 counts and targeted HIV viral load (VL) to confirm drug failure. VL measurements and HIVDR genotyping were performed retrospectively on baseline and month 12 samples. One hundred and fifty-eight participants who completed their month 12 follow-up visit had VL data available at month 12. Most of them (88%) were virologically suppressed (VL≤1000 copies/mL) but 18 had virological failure (11%), which is in the range of WHO-suggested targets for HIVDR prevention. If only CD4 criteria had been used to classify treatment response, 26% of the participants would have been misclassified as treatment failure. Pre-therapy HIVDR was documented in 4 of 109 participants (3.6%) with an HIVDR genotyping results at baseline. Eight of 12 participants (66.7%) with virological failure and HIVDR genotyping results at month 12 were found to harbor mutation(s), mostly NNRTI resistance mutations, whereas 4 patients had no HIVDR mutations. Almost half (44%) of the participants initiated ART at CD4 count ≤200cell/µl and severe CD4 depletion at baseline (<50 cells/µl) was associated with virological treatment failure (p = 0.008).

Although the findings may not be generalizable to all HIV patients in Rwanda, our data suggest that first-line ART regimen changes are currently not warranted. However, the accumulation of acquired HIVDR mutations in some participants underscores the need to reinforce HIVDR prevention strategies, such as increasing the availability and appropriate use of VL testing to monitor ART response, ensuring high quality adherence counseling, and promoting earlier identification of HIV patients and enrollment into HIV care and treatment programs.

Association of C-reactive protein and interleukin-6 with new-onset fatigue in the Whitehall II prospective cohort study

BACKGROUND

Although basic research on neuroimmune interactions suggests that inflammatory processes may play a role in the development of fatigue, population-based evidence on this association is limited. This study examined whether plasma C-reactive protein (CRP) and interleukin-6 (IL-6), biomarkers of systemic inflammation, predict fatigue onset.

METHOD

The Whitehall II study is a large-scale cohort study conducted in 20 civil service departments in London. Plasma CRP and IL-6 were measured in 4847 non-fatigued participants at phase 3 (1991-1993, aged 39-63 years). Fatigue was assessed using the Vitality subscale of the 36-item Short Form Health Survey (SF-36) at phase 3 and phase 4 (1995-1996).

RESULTS

During a mean follow-up of 3.1 years, 957 new fatigue cases (19.7%) were identified using the pre-established cut-off score of ≤ 50 on the Vitality subscale. CRP values were dichotomized as low (<1.0 mg/l ) or high (≥ 1.0 mg/l) using the Centers for Disease Control/American Heart Association recommendations. Similarly, IL-6 values were also dichotomized as low (<1.5 pg/ml) or high (≥ 1.5 pg/ml). After full adjustment for sociodemographic and biobehavioral covariates, the odds ratios for new-onset fatigue were 1.28 [95% confidence interval (CI) 1.09-1.49, p = 0.003] for high CRP and 1.24 (95% CI 1.06-1.45, p = 0.008) for high IL-6. Similar results were found when CRP and IL-6 were treated as continuous variables.

CONCLUSIONS

Plasma CRP and IL-6 were prospectively associated with new-onset fatigue, supporting the hypothesis that low-grade inflammation has a role in the development of fatigue.

Logistic regression for risk factor modelling in stuttering research

OBJECTIVES

To outline the uses of logistic regression and other statistical methods for risk factor analysis in the context of research on stuttering.

DESIGN

The principles underlying the application of a logistic regression are illustrated, and the types of questions to which such a technique has been applied in the stuttering field are outlined. The assumptions and limitations of the technique are discussed with respect to existing stuttering research, and with respect to formulating appropriate research strategies to accommodate these considerations. Finally, some alternatives to the approach are briefly discussed.

RESULTS

The way the statistical procedures are employed are demonstrated with some hypothetical data.

CONCLUSION

Research into several practical issues concerning stuttering could benefit if risk factor modelling were used. Important examples are early diagnosis, prognosis (whether a child will recover or persist) and assessment of treatment outcome.

EDUCATIONAL OBJECTIVES

After reading this article you will: (a) Summarize the situations in which logistic regression can be applied to a range of issues about stuttering; (b) Follow the steps in performing a logistic regression analysis; (c) Describe the assumptions of the logistic regression technique and the precautions that need to be checked when it is employed; (d) Be able to summarize its advantages over other techniques like estimation of group differences and simple regression.

Diagnostic and prognostic prediction models

Risk prediction models can be used to estimate the probability of either having (diagnostic model) or developing a particular disease or outcome (prognostic model). In clinical practice, these models are used to inform patients and guide therapeutic management. Examples from the field of venous thrombo-embolism (VTE) include the Wells rule for patients suspected of deep venous thrombosis and pulmonary embolism, and more recently prediction rules to estimate the risk of recurrence after a first episode of unprovoked VTE. In this paper, the three phases that are recommended before a prediction model may be used in daily practice are described: development, validation, and impact assessment. In the development phase, the focus is on model development commonly using a multivariable logistic (diagnostic) or survival (prognostic) regression analysis. The performance of the developed model is expressed by discrimination, calibration and (re-) classification. In the validation phase, the developed model is tested in a new set of patients using these same performance measures. This is important, as model performance is commonly poorer in a new set of patients, e.g. due to case-mix or domain differences. Finally, in the impact phase the ability of a prediction model to actually guide patient management is evaluated. Whereas in the development and validation phase single cohort designs are preferred, this last phase asks for comparative designs, ideally randomized designs; therapeutic management and outcomes after using the prediction model is compared to a control group not using the model (e.g. usual care).

Imaging indexes of activity and severity for Crohn's disease: current status and future trends

BACKGROUND

Cross-sectional imaging techniques, including ultrasonography (US), computed tomography (CT), and magnetic resonance imaging (MRI) are increasingly used for evaluation of Crohn's disease (CD).

AIM

To review the accuracy of cross-sectional imaging indexes for measurement of disease activity and severity in patients with CD, and to evaluate its applicability on daily clinical practice and research.

METHODS

Relevant publications were identified by literature search, and selected based on predefined quality parameters, including a sample size and reference standard. Nineteen publications were chosen.

RESULTS

The US-based indexes of activity showed high correlation with reference standard indexes. There is a good or very good agreement between the MR-based indexes and the reference standard when the comparison is limited to small segments of intestine. Significant discrepancies have been found between indexes that evaluate the colon. Only one CT-based index was included. The main strengths and weaknesses of the indexes, according to its design, are discussed.

CONCLUSION

Standardization of image acquisition protocols and patient preparation should be procured, especially for MRI. In daily practice, a simple, semi-quantitative index providing relevant information on disease activity and severity is preferable. For research purposes, a precise and reproducible index should be mandatory.

A method for estimating the risk of surgical site infection in patients with abdominal colorectal procedures

BACKGROUND

Surgical site infection is one of the most common and significant morbidities following colon and rectal surgery, representing a marker of institutional quality. Various measures have been implemented to lower its incidence. However, the level of incidence remains unacceptable in many reports.

OBJECTIVE

This study addresses whether surgical site infections can be accurately predicted in an outpatient clinical setting among patients undergoing elective colon and rectal surgery.

DESIGN

This investigation was designed as a retrospective cohort study with the use of logistic regression modeling.

SETTINGS

Data for this study were extracted from the American College of Surgeons National Surgical Quality Improvement Program Participant user data file.

PATIENTS

Patients undergoing elective intraabdominal colorectal surgery during 2009 were included.

MAIN OUTCOME MEASURES

The primary outcome measured was the probability of 30-day surgical site infection (superficial and deep incisional).

RESULTS

A total of 18,403 records for patients with colorectal surgery were identified. Superficial incisional surgical site infections were identified in 1447 records (7.86%). Deep incisional surgical site infections were identified in 278 records (1.51%). Body mass index, preoperative hematocrit, open approach, ASA classification level, smoking, alcohol use, functional status before surgery, and age more than 75 years were identified as likely independent predictors of deep and superficial surgical site infections. Multivariable logistic regression analysis was used to develop a series of predictive models. Reduced versions of the models were then developed that included only highly statistically significant predictors of infection in the corresponding full models (age, alcohol abuse, ASA classification, stoma closure, open approach, BMI, and hematocrit). Nomograms representing the final reduced model equations are presented.

LIMITATIONS

This study was limited by the use of an administrative database and its retrospective design.

CONCLUSIONS

Surgical site infection is common morbidity following colon and rectal surgery. Nomograms using key patient characteristics can be used to accurately calculate a patients' risk of surgical site infection. This tool could be applied in the clinical setting to prospectively identify patients at highest risk of surgical site infection.

Risk factors for falls in multiple sclerosis: an observational study

Background:

People with multiple sclerosis (MS) experience frequent falls, which are associated with impairments and limitations to activities and participations.

Objective:

The objective of this paper is to evaluate falls risk factors using robust clinical measures.

Methods:

A total of 150 individuals (confirmed MS diagnosis, Expanded Disability Status Scale (EDSS) 3.5–6.5) were recruited, with 148 participants included in the final analysis. Demographic data were collected and performance assessed in eight predictor measures (Physiological Profile Assessment (PPA), Brief Ataxia Rating scale, Ashworth scale (Ashworth), Modified Falls Efficacy scale, Symbol Digit Modalities Test, dual-task interference, lying/standing blood pressure, static/dynamic visual acuity). Participants prospectively recorded falls over three months using a daily diary. People were classified as “fallers” based on reports of ≥ two falls.

Results:

A total of 104 participants recorded 672 falls; 78 (52.7%) reported ≥ two falls. Continence issues, previous falls history and use of prescribed medications were each associated with increased risk of being a “faller”. Ashworth and PPA risk score contributed significantly to a logistic regression model predicting faller/non-faller classification. The reduced model (Ashworth, PPA, EDSS) showed fair-to-good predictive ability (ROC c-statistic 0.73, sensitivity 70%, specificity 69%).

Conclusion:

This study confirms the high prevalence of falls in ambulant people with MS. Important potentially modifiable risk factors are identified, suggesting aspects to target in falls interventions.

When should a DDH experiment be mandatory in microbial taxonomy?

DNA-DNA hybridizations (DDH) play a key role in microbial species discrimination in cases when 16S rRNA gene sequence similarities are 97 % or higher. Using real-world 16S rRNA gene sequences and DDH data, we here re-investigate whether or not, and in which situations, this threshold value might be too conservative. Statistical estimates of these thresholds are calculated in general as well as more specifically for a number of phyla that are frequently subjected to DDH. Among several methods to infer 16S gene sequence similarities investigated, most of those routinely applied by taxonomists appear well suited for the task. The effects of using distinct DDH methods also seem to be insignificant. Depending on the investigated taxonomic group, a threshold between 98.2 and 99.0 % appears reasonable. In that way, up to half of the currently conducted DDH experiments could safely be omitted without a significant risk for wrongly differentiated species.

Clinical prediction model to aid emergency doctors managing febrile children at risk of serious bacterial infections: diagnostic study

OBJECTIVE

To derive, cross validate, and externally validate a clinical prediction model that assesses the risks of different serious bacterial infections in children with fever at the emergency department.

DESIGN

Prospective observational diagnostic study.

SETTING

Three paediatric emergency care units: two in the Netherlands and one in the United Kingdom.

PARTICIPANTS

Children with fever, aged 1 month to 15 years, at three paediatric emergency care units: Rotterdam (n=1750) and the Hague (n=967), the Netherlands, and Coventry (n=487), United Kingdom. A prediction model was constructed using multivariable polytomous logistic regression analysis and included the predefined predictor variables age, duration of fever, tachycardia, temperature, tachypnoea, ill appearance, chest wall retractions, prolonged capillary refill time (>3 seconds), oxygen saturation <94%, and C reactive protein.

MAIN OUTCOME MEASURES

Pneumonia, other serious bacterial infections (SBIs, including septicaemia/meningitis, urinary tract infections, and others), and no SBIs.

RESULTS

Oxygen saturation <94% and presence of tachypnoea were important predictors of pneumonia. A raised C reactive protein level predicted the presence of both pneumonia and other SBIs, whereas chest wall retractions and oxygen saturation <94% were useful to rule out the presence of other SBIs. Discriminative ability (C statistic) to predict pneumonia was 0.81 (95% confidence interval 0.73 to 0.88); for other SBIs this was even better: 0.86 (0.79 to 0.92). Risk thresholds of 10% or more were useful to identify children with serious bacterial infections; risk thresholds less than 2.5% were useful to rule out the presence of serious bacterial infections. External validation showed good discrimination for the prediction of pneumonia (0.81, 0.69 to 0.93); discriminative ability for the prediction of other SBIs was lower (0.69, 0.53 to 0.86).

CONCLUSION

A validated prediction model, including clinical signs, symptoms, and C reactive protein level, was useful for estimating the likelihood of pneumonia and other SBIs in children with fever, such as septicaemia/meningitis and urinary tract infections.

Maternal plasma concentrations of angiogenic/antiangiogenic factors in the third trimester of pregnancy to identify the patient at risk for stillbirth at or near term and severe late preeclampsia

OBJECTIVE

To determine whether maternal plasma concentrations of placental growth factor (PlGF), soluble endoglin (sEng), and soluble vascular endothelial growth factor receptor-1 (sVEGFR-1) at 30-34 weeks of gestation can identify patients at risk for stillbirth, late preeclampsia, and delivery of small-for-gestational-age (SGA) neonates.

STUDY DESIGN

A prospective cohort study included 1269 singleton pregnant women from whom blood samples were obtained at 30-34 weeks of gestation and who delivered at >34 weeks of gestation. Plasma concentrations of PlGF, sEng, and sVEGFR-1 were determined by enzyme-linked immunosorbent assay.

RESULTS

The prevalence of late (>34 weeks of gestation) preeclampsia, severe late preeclampsia, stillbirth, and SGA was 3.2% (n = 40), 1.8% (n = 23), 0.4% (n = 5), and 8.5% (n = 108), respectively. A plasma concentration of PlGF/sEng <0.3 MoM was associated with severe late preeclampsia (adjusted odds ratio, 16); the addition of PlGF/sEng to clinical risk factors increased the area under the receiver-operating characteristic curve from 0.76 to 0.88 (P = .03). The ratio of PlGF/sEng or PlGF/sVEGFR-1 in the third trimester outperformed those obtained in the first or second trimester and uterine artery Doppler velocimetry at 20-25 weeks of gestation for the prediction of severe late preeclampsia (comparison of areas under the receiver-operating characteristic curve; each P ≤ .02). Both PlGF/sEng and PlGF/sVEGFR-1 ratios achieved a sensitivity of 74% with a fixed false-positive rate of 15% for the identification of severe late preeclampsia. A plasma concentration of PlGF/sVEGFR-1 <0.12 MoM at 30-34 weeks of gestation had a sensitivity of 80%, a specificity of 94%, and a likelihood ratio of a positive test of 14 for the identification of subsequent stillbirth. Similar findings (sensitivity 80%; specificity 93%) were observed in a separate case-control study.

CONCLUSION

Risk assessment for stillbirth and severe late preeclampsia in the third trimester is possible with the determination of maternal plasma concentrations of angiogenic and antiangiogenic factors at 30-34 weeks of gestation.

Donor-recipient matching: myths and realities

Liver transplant outcomes keep improving, with refinements of surgical technique, immunosuppression and post-transplant care. However, these excellent results and the limited number of organs available have led to an increasing number of potential recipients with end-stage liver disease worldwide. Deaths on waiting lists have led liver transplant teams maximize every organ offered and used in terms of pre and post-transplant benefit. Donor-recipient (D-R) matching could be defined as the technique to check D-R pairs adequately associated by the presence of the constituents of some patterns from donor and patient variables. D-R matching has been strongly analysed and policies in donor allocation have tried to maximize organ utilization whilst still protecting individual interests. However, D-R matching has been written through trial and error and the development of each new score has been followed by strong discrepancies and controversies. Current allocation systems are based on isolated or combined donor or recipient characteristics. This review intends to analyze current knowledge about D-R matching methods, focusing on three main categories: patient-based policies, donor-based policies and combined donor-recipient systems. All of them lay on three mainstays that support three different concepts of D-R matching: prioritarianism (favouring the worst-off), utilitarianism (maximising total benefit) and social benefit (cost-effectiveness). All of them, with their pros and cons, offer an exciting controversial topic to be discussed. All of them together define D-R matching today, turning into myth what we considered a reality in the past.

Evidence of an imbalance of angiogenic/antiangiogenic factors in massive perivillous fibrin deposition (maternal floor infarction): a placental lesion associated with recurrent miscarriage and fetal death

OBJECTIVE

Massive perivillous fibrin deposition (MPFD) is associated with serious complications of pregnancy including recurrent spontaneous abortion, fetal growth restriction, and fetal demise. The aim of this study was to determine whether maternal plasma concentrations of angiogenic/antiangiogenic factors in MPFD differ from those of uncomplicated pregnancies.

STUDY DESIGN

This retrospective longitudinal case-control study included MPFD cases (n = 10) and control patients (n = 175) with uncomplicated pregnancies who were enrolled in a longitudinal study and delivered at term. Serial plasma concentrations of placental growth factor (PlGF), soluble endoglin (sEng), and soluble vascular endothelial growth factor receptor (sVEGFR)-1 and -2 were determined by an enzyme-linked immunosorbent assay (cases, n = 28 samples; controls, n = 723 samples). Individual analyte concentrations were averaged across gestational age at specimen collection intervals. Linear mixed models were used to test for differences in log-transformed mean analyte concentrations both overall and as a function of time.

RESULTS

The following results were found: (1) patients with MPFD had a lower mean plasma PlGF concentration (P = .03) and higher mean plasma concentrations of sVEGFR-1 and sEng (both P < .01) than controls, adjusted for potential confounders; (2) the mean plasma concentration of PlGF differed further among cases and controls as a function of gestational age interval (P < .0001); however, mean sVEGFR-1 and sEng group differences as a function of gestational age interval approached but did not reach significance (P = .09 and P = .11, respectively); (3) patients with MPFD had lower mean plasma concentrations of PlGF/sVEGFR-1 (P < .0001) and PlGF/sEng (P < .001): both of these relationships differed further as a function of gestational age interval (both P < .0001); and (4) differences in mean sVEGFR-1, sEng, and the ratios of PlGF to sVEGFR-1 and PlGF to sEng were observed before 20 weeks of gestation.

CONCLUSION

An imbalance of angiogenic/antiangiogenic factors is present in patients with MPFD prior to the diagnosis. We propose that these changes participate in the mechanisms responsible for adverse pregnancy outcomes in patients with MPFD.

Dutch postgraduate GP selection procedure; reliability of interview assessments

Background

Semi- structured interviews are the core of the Dutch selection procedure for postgraduate general practice (GP) training. A staff member, trainer and trainee independently assess personal qualities. Aiming to improve the selection procedure we were interested in the reliability aspects of these interviews. We investigated the inter-rater reliability of the interview for groups of two or three assessors and the degree to which candidates’ characteristics and qualities assessed during interviews explained admission into GP training, controlled for differences between those who apply for the first versus the second or third application.

Methods

An observational study was conducted of all candidates who entered the Utrecht selection procedure between April 2008 and 2010. Candidates’ characteristics and qualities were collected. Inter-rater reliability of different compositions of the interview group per quality was estimated. Factors associated with admission into GP training were assessed.

Results

The study population included 394 candidates. Twenty-six candidates were rejected based on their application letter (4.4%). Three candidates who applied more than 3 times were excluded. Ultimately, 206 of the 365 candidates were admitted to the GP training (56,4%). The inter-rater reliability was satisfactory (ICC: 0.78 – 0 .84). Reduction from three to two assessors slightly reduces the ICC. The candidates’ qualities independently explained admission to GP training, whereas individual characteristics did not. These results did not differ for candidates who applied for the first time versus candidates applying for the second or third time.

Conclusion

Selection interviews with two assessors yielded a satisfactory level of reliability. Individual characteristics were not associated with admission, whereas scores related to candidate qualities did show such an association. The results of those applying for the second or third time were similar.

Improving the performance of risk-adjusted mortality modeling for colorectal cancer surgery by combining claims data and clinical data

OBJECTIVES

The objective of this study was to evaluate the performance of risk-adjusted mortality models for colorectal cancer surgery.

METHODS

We investigated patients (n=652) who had undergone colorectal cancer surgery (colectomy, colectomy of the rectum and sigmoid colon, total colectomy, total proctectomy) at five teaching hospitals during 2008. Mortality was defined as 30-day or in-hospital surgical mortality. Risk-adjusted mortality models were constructed using claims data (basic model) with the addition of TNM staging (TNM model), physiological data (physiological model), surgical data (surgical model), or all clinical data (composite model). Multiple logistic regression analysis was performed to develop the risk-adjustment models. To compare the performance of the models, both c-statistics using Hanley-McNeil pair-wise testing and the ratio of the observed to the expected mortality within quartiles of mortality risk were evaluated to assess the abilities of discrimination and calibration.

RESULTS

The physiological model (c=0.92), surgical model (c=0.92), and composite model (c=0.93) displayed a similar improvement in discrimination, whereas the TNM model (c=0.87) displayed little improvement over the basic model (c=0.86). The discriminatory power of the models did not differ by the Hanley-McNeil test (p>0.05). Within each quartile of mortality, the composite and surgical models displayed an expected mortality ratio close to 1.

CONCLUSIONS

The addition of clinical data to claims data efficiently enhances the performance of the risk-adjusted postoperative mortality models in colorectal cancer surgery. We recommended that the performance of models should be evaluated through both discrimination and calibration.

Differences between HIV-positive and HIV-negative African American men who have sex with men in two major U.S. metropolitan areas

Alarmingly high HIV prevalence rates among African American men who have sex with men (AAMSM) require the development of effective prevention interventions. In this study of AAMSM conducted in two cities, we explored similarities and differences between HIV-positive and HIV-negative AAMSM on sociodemographic variables, HIV-related risk behaviors, and attitudinal constructs. Differences emerged in several major life areas: (1) poverty, employment, and use of mental health services, (2) sexual risk behaviors, and (3) self-identification with gay identity and culture. With regard to sociodemographic indicators, HIV-positive AAMSM were doing worse than HIV-negative AAMSM in that they were more likely to be disabled, to be living below the poverty level, and accessing mental health services. With regard to risk behaviors and partner characteristics, HIV-positive AAMSM were acting more responsibly than their HIV-negative counterparts, as they were more likely to have used a condom the last time they had sex. In addition, when compared to their HIV-negative counterparts, HIV-positive AAMSM were more likely to have either no casual partners at all or main or casual partners who were HIV-positive, thus preventing new HIV transmission by partnering with other HIV-positive men. Attitudinally, HIV-positive men were more accepting of their sexual attractions to men and were more likely to identify as gay than their HIV-negative peers. Although causality cannot be determined, the findings of this study can be used to strengthen HIV prevention efforts by improving the selection of targeted behaviors and prevention messages for HIV-positive and HIV-negative AAMSM.

Vascular closure device failure in contemporary practice

OBJECTIVES

The goal of this study was to assess the frequency and predictors of vascular closure device (VCD) deployment failure, and its association with vascular complications of 3 commonly used VCDs.

BACKGROUND

VCDs are commonly used following percutaneous coronary intervention on the basis of studies demonstrating reduced time to ambulation, increased patient comfort, and possible reduction in vascular complications as compared with manual compression. However, limited data are available on the frequency and predictors of VCD failure, and the association of deployment failure with vascular complications.

METHODS

From a de-identified dataset provided by Massachusetts Department of Health, 23,813 consecutive interventional coronary procedures that used either a collagen plug-based (n = 18,533), a nitinol clip-based (n = 2,284), or a suture-based (n = 2,996) VCD between June 2005 and December 2007 were identified. The authors defined VCD failure as unsuccessful deployment or failure to achieve immediate access site hemostasis.

RESULTS

Among 23,813 procedures, the VCD failed in 781 (3.3%) procedures (2.1% of collagen plug-based, 6.1% of suture-based, 9.5% of nitinol clip-based VCDs). Patients with VCD failure had an excess risk of "any" (7.7% vs. 2.8%; p < 0.001), major (3.3% vs. 0.8%; p < 0.001), or minor (5.8% vs. 2.1%; p < 0.001) vascular complications compared with successful VCD deployment. In a propensity score-adjusted analysis, when compared with collagen plug-based VCD (reference odds ratio [OR] = 1.0), nitinol clip-based VCD had 2-fold increased risk (OR: 2.0, 95% confidence interval [CI]: 1.8 to 2.3, p < 0.001) and suture-based VCD had 1.25-fold increased risk (OR: 1.25, 95% CI: 1.2 to 1.3, p < 0.001) for VCD failure. VCD failure was a significant predictor of subsequent vascular complications for both collagen plug-based VCD and nitinol clip-based VCD, but not for suture-based VCD.

CONCLUSIONS

VCD failure rates vary depending upon the type of VCD used and are associated with significantly higher vascular complications as compared with deployment successes.

Phytotherapeutic management of polycystic ovary syndrome: role of aerial parts of wood betony (Stachys lavandulifolia)

Pharmacological treatments of polycystic ovary syndrome (PCOS) have side effects. We compared the efficacy and safety of the aerial parts of wood botany (AWB) with those of Medroxyprogesterone acetate (MPA) in the management of abnormal uterine bleeding (AUB) due to PCOS. We conducted a randomized clinical trial of 66 women aged 15-45 years, with AUB. Participants were randomly assigned to either cyclical 10-mg BD MPA for three cycles or 5 g ABW TDS for three months. Clinical symptoms and paraclinical parameters were recorded before and 3 months after initiation of the treatment. Comparisons were made using generalized linear models. Age-adjusted prevalence rate of different patterns of AUB, decreased from 2.7 (95%CIs: 0.6-48.0) to 1.1 (95%CIs: 0.1-18.3) for patients taking MPA, and from 2.5 (95% CIs: 0.2-40.2) to 0.7 (95%CIs: 0.0-12.1). Decrease in prevalence rate was similar across two arms of the study (P value = 0.248). Adverse effects were observed less frequently (24.2%) among participants on MPA than among those on AWB (45.5%). The multivariate-adjusted odds for developing adverse reaction of MPA was 0.40 (95%CIs: 0.14-1.19, P value = 0.099) time odds of AWB. In conclusion, AWB may be used as an alternative for MPA in the treatment of AUB caused by PCOS.

Transient hypothyroidism at 3-year follow-up among cases of congenital hypothyroidism detected by newborn screening

OBJECTIVE

To investigate the rate of transient thyroid deficiency and treatment compliance among cases with congenital hypothyroidism diagnosed and followed-up after age 3 years by newborn screening (NBS).

STUDY DESIGN

Cases detected by Michigan NBS between October 1, 2003, and December 31, 2007, and followed-up after age 3 years were included. The χ(2) and Fisher exact tests were used to test differences among followed and lost cases. Logistic regression models were used to investigate predictors of treatment cessation.

RESULTS

Roughly 45% of eligible cases were lost to follow-up, and disease state (transient or permanent congenital hypothyroidism) could not be determined for 12 cases (7.9%). Of the 72 followed cases, 34 (47%) were considered permanent congenital hypothyroidism based on thyroid imaging findings (n = 7) or an increase in medication dosage over time (n = 27). One-quarter of followed cases with congenital hypothyroidism were no longer being treated, and of these, just over 83% stopped treatment without medical supervision. Of 23 cases that underwent a medically supervised trial without thyroid hormone medication, treatment was reinstated in 20. Laboratory confirmation of euthyroidism was available for 6 of 18 cases clinically deemed transient. After adjustment, black race was the strongest predictor of treatment cessation (OR, 9.86; 95% CI, 1.82-53.31). Treatment cessation was also more common among low birth weight infants and those admitted to the neonatal intensive care unit at birth.

CONCLUSION

We recommend that NBS programs include long-term follow-up through at least age 3 years to determine treatment compliance and disease permanence. Further research is needed to determine ideal follow-up program operations and reassessment methods for congenital hypothyroidism disease permanence. Guidelines that provide evidence-based reassessment methods would be beneficial for the healthcare providers of children with congenital hypothyroidism.

Cardiac operative risk evaluation: The EuroSCORE II, does it make a real difference?

BACKGROUND

The EuroSCORE, worldwide used as a model for prediction of mortality after cardiac surgery, has recently been renewed. Since October 2011, the EuroSCORE II calculator is available at the EuroSCORE website and recommended for clinical use. The intention of this paper is to compare the use of the initial EuroSCORE and EuroSCORE II as a risk evaluation tool.

METHODS

100 consecutive patients who underwent combined mitral valve and coronary bypass surgery (MVR + CABG) and 100 consecutive patients undergoing combined aortic valve surgery and coronary bypass surgery (AVR + CABG) at the Radboud University Nijmegen Medical Center before 10 October 2011 were included. For both groups the initial EuroSCORE and the EuroSCORE II model were used for risk calculation and based on the calculated risks, cumulative sum charts (CUSUM) were constructed to evaluate the impact on performance monitoring.

RESULTS

For the MVR + CABG group the calculated risk using the initial logistic EuroSCORE was 9.95 ± 8.47 (1.51-45.37) versus 5.08 ± 4.03 (0.67-19.76) for the EuroSCORE II. For the AVR + CABG group 9.50 ± 8.6 (1.51-69.5) versus 4.77 ± 6.6 (0.96-64.24), respectively. For both groups the calculated risk by the EuroSCORE II was statistically lower compared with the initial EuroSCORE (p < 0.001). This lower expected risk has influence on performance monitoring, using risk-adjusted CUSUM analysis.

CONCLUSION

The EuroSCORE II, based on a recently updated database, reduces the overestimation of the calculated risk by the initial EuroSCORE. This difference is statistically significant and the EuroSCORE II may also reflect better current surgical performance.

Impact of follicular G-CSF quantification on subsequent embryo transfer decisions: a proof of concept study

BACKGROUND

Previous experiments have shown that granulocyte colony-stimulating factor (G-CSF), quantified in the follicular fluid (FF) of individual oocytes, correlates with the potential for an ongoing pregnancy of the corresponding fertilized oocytes among selected transferred embryos. Here we present a proof of concept study aimed at evaluating the impact of including FF G-CSF quantification in the embryo transfer decisions.

METHODS

FF G-CSF was quantified with the Luminex XMap technology in 523 individual FF samples corresponding to 116 fresh transferred embryos, 275 frozen embryos and 131 destroyed embryos from 78 patients undergoing ICSI.

RESULTS

Follicular G-CSF was highly predictive of subsequent implantation. The receiving operator characteristics curve methodology showed its higher discriminatory power to predict ongoing pregnancy in multivariate logistic regression analysis for FF G-CSF compared with embryo morphology [0.77 (0.69–0.83), P < 0.001 versus 0.66 (0.58–0.73), P = 0.01)]. Embryos were classified by their FF G-CSF concentration: Class I over 30 pg/ml (a highest positive predictive value for implantation), Class II from 30 to 18.4 pg/ml and Class III <18.4 pg/ml (a highest negative predictive value). Embryos derived from Class I follicles had a significantly higher implantation rate (IR) than those from Class II and III follicles (36 versus 16.6 and 6%, P < 0.001). Embryos derived from Class I follicles with an optimal morphology reached an IR of 54%. Frozen-thawed embryos transfer derived from Class I follicles had an IR of 37% significantly higher than those from Class II and III follicles, respectively, of 8 and 5% (P < 0.001). Thirty-five per cent of the frozen embryos but also 10% of the destroyed embryos were derived from G-CSF Class I follicles. Non-optimal embryos appear to have been transferred in 28% (22/78) of the women, and their pregnancy rate was significantly lower than that of women who received at least one optimal embryo (18 versus 36%, P = 0.04).

CONCLUSIONS

Monitoring FF G-CSF for the selection of embryos with a better potential for pregnancy might improve the effectiveness of IVF by reducing the time and cost required for obtaining a pregnancy.

Risk stratification in acute heart failure: rationale and design of the STRATIFY and DECIDE studies

BACKGROUND

A critical challenge for physicians facing patients presenting with signs and symptoms of acute heart failure (AHF) is how and where to best manage them. Currently, most patients evaluated for AHF are admitted to the hospital, yet not all warrant inpatient care. Up to 50% of admissions could be potentially avoided and many admitted patients could be discharged after a short period of observation and treatment. Methods for identifying patients that can be sent home early are lacking. Improving the physician's ability to identify and safely manage low-risk patients is essential to avoiding unnecessary use of hospital beds.

METHODS

Two studies (STRATIFY and DECIDE) have been funded by the National Heart Lung and Blood Institute with the goal of developing prediction rules to facilitate early decision making in AHF. Using prospectively gathered evaluation and treatment data from the acute setting (STRATIFY) and early inpatient stay (DECIDE), rules will be generated to predict risk for death and serious complications. Subsequent studies will be designed to test the external validity, utility, generalizability and cost-effectiveness of these prediction rules in different acute care environments representing racially and socioeconomically diverse patient populations.

RESULTS

A major innovation is prediction of 5-day as well as 30-day outcomes, overcoming the limitation that 30-day outcomes are highly dependent on unpredictable, post-visit patient and provider behavior. A novel aspect of the proposed project is the use of a comprehensive cardiology review to correctly assign post-treatment outcomes to the acute presentation.

CONCLUSIONS

Finally, a rigorous analysis plan has been developed to construct the prediction rules that will maximally extract both the statistical and clinical properties of every data element. Upon completion of this study we will subsequently externally test the prediction rules in a heterogeneous patient cohort.

The risky business of studying prognosis

Prognosis studies provide important healthcare information. Clinicians use prognostic factors to predict disease progress, thus allowing individualization of disease management. Prognosis is the issue in many translational studies that aim to identify biomarkers to predict outcomes. In a clinical trial, researchers may use prognostic factors to sort patients into risk groups, to clarify the effects of a new therapeutic agent. Prognosis studies can have significant effects on clinical practice.