Logistic regression modeling and the number of events per variable: selection bias dominates

Nivolumab and ipilimumab in recurrent or refractory cancer of unknown primary: a phase II trial

Cancer of unknown primary has a dismal prognosis, especially following failure of platinum-based chemotherapy. 10-20% of patients have a high tumor mutational burden (TMB), which predicts response to immunotherapy in many cancer types. In this prospective, non-randomized, open-label, multicenter Phase II trial (EudraCT 2018-004562-33; NCT04131621), patients relapsed or refractory after platinum-based chemotherapy received nivolumab and ipilimumab following TMBhigh vs. TMBlow stratification. Progression-free survival (PFS) represented the primary endpoint; overall survival (OS), response rates, duration of clinical benefit and safety were the secondary endpoints. The trial was prematurely terminated in March 2021 before reaching the preplanned sample size (n = 194). Among 31 evaluable patients, 16% had a high TMB ( > 12 mutations/Mb). Overall response rate was 16% (95% CI 6-34%), with 7.7% (95% CI 1-25%) vs. 60% (95% CI 15-95%) in TMBlow and TMBhigh, respectively. Although the primary endpoint was not met, high TMB was associated with better median PFS (18.3 vs. 2.4 months) and OS (18.3 vs. 3.6 months). Severe immune-related adverse events were reported in 29% of cases. Assessing on-treatment dynamics of circulating tumor DNA using combined targeted hotspot mutation and shallow whole genome sequencing as part of a predefined exploratory analysis identified patients benefiting from immunotherapy irrespective of initial radiologic response.

Survivorship analysis of revision reverse total shoulder arthroplasty

BACKGROUND

The expansion of indications for reverse total shoulder arthroplasty (RTSA) has resulted in a rapid increase in the incidence of subsequent revision procedures. The purpose of this study was to identify the incidence and risk factors for re-revision shoulder arthroplasty after first revision RTSA.

METHODS

We retrospectively queried our institutional shoulder arthroplasty database of prospectively collected data from 2003 to 2019. To assess revision implant survival, patients were censored on the date of re-revision surgery or, if the revision arthroplasty was not revised, at the most recent follow-up or their date of death. Patients with a prior infection, concern for infection at the time of revision, antibiotic spacer, or oncologic indication for primary arthroplasty were excluded. A total of 186 revision RTSAs were included, with 32 undergoing re-revision shoulder arthroplasty. The Kaplan-Meier method and bivariate Cox regression were used to assess the relationship of patient and surgical characteristics on implant survivorship. Multivariate Cox regression was performed to identify independent predictors of re-revision.

RESULTS

Re-revision shoulder arthroplasty was most commonly performed for instability (34%), infection (28%), and glenoid loosening (19%). Overall re-revision rates at 6 months (7%), 1 year (9%), and 2 years (13%) were relatively low; however, the rate of re-revision increased at 5 years (35%). Men underwent re-revision more often than women within the first 6 months after revision RTSA (12% vs. 2%; P = .025), but not thereafter. On multivariate analysis, increased estimated blood loss was associated with a greater risk of undergoing re-revision shoulder arthroplasty (hazard ratio: 41.16 [3.34-506.50]; P = .004).

CONCLUSION

The rate of re-revision after revision RTSA is low in the first 2 years postoperatively (13%) but increases to 35% at 5 years. Increased estimated blood loss, which may reflect greater operative complexity, was identified as a risk factor that may confer an increased chance of re-revision after revision RTSA. Knowledge of risk factors for re-revision after revision RTSA can aid surgeons and patients in preoperative counseling.

Association of Racial Disparities With In-Hospital Outcomes in Severe Bronchopulmonary Dysplasia

IMPORTANCE

Bronchopulmonary dysplasia (BPD) is the most common serious morbidity of preterm birth. Short-term respiratory outcomes for infants with the most severe forms of BPD are highly variable. The mechanisms that explain this variability remain unknown and may be mediated by racial disparities.

OBJECTIVE

To determine the association of maternal race with death and length of hospital stay in a multicenter cohort of infants with severe BPD.

DESIGN, SETTING, AND PARTICIPANTS

This multicenter cohort study included preterm infants enrolled in the BPD Collaborative registry from January 1, 2015, to July 19, 2021, involving 8 BPD Collaborative centers located in the US. Included patients were born at less than 32 weeks' gestation, had a diagnosis of severe BPD as defined by the 2001 National Institutes of Health Consensus Criteria, and were born to Black or White mothers.

EXPOSURES

Maternal race: Black vs White.

MAIN OUTCOMES AND MEASURES

Death and length of hospital stay.

RESULTS

Among 834 registry infants (median [IQR] gestational age, 25 [24-27] weeks; 492 male infants [59%]) meeting inclusion criteria, the majority were born to White mothers (558 [67%]). Death was observed infrequently in the study cohort (32 [4%]), but Black maternal race was associated with an increased odds of death (adjusted odds ratio, 2.1; 95% CI, 1.2-3.5) after adjusting for center. Black maternal race was also significantly associated with length of hospital stay (adjusted between-group difference, 10 days; 95% CI, 3-17 days).

CONCLUSIONS AND RELEVANCE

In a multicenter severe BPD cohort, study results suggest that infants born to Black mothers had increased likelihood of death and increased length of hospital stay compared with infants born to White mothers. Prospective studies are needed to define the sociodemographic mechanisms underlying disparate health outcomes for Black infants with severe BPD.

Comparative Analyses of the Impact of Different Criteria for Sepsis Diagnosis on Outcome in Patients with Spontaneous Subarachnoid Hemorrhage

Data on sepsis in patients with a subarachnoid hemorrhage (SAH) are scarce. We assessed the impact of different sepsis criteria on the outcome in an SAH cohort. Adult patients admitted to our ICU with a spontaneous SAH between 11/2014 and 11/2018 were retrospectively included. In patients developing an infection, different criteria for sepsis diagnosis (Sepsis-1, Sepsis-3_original, Sepsis-3_modified accounting for SAH-specific therapy, alternative sepsis criteria compiled of consensus conferences) were applied and their impact on functional outcome using the modified Rankin Scale (mRS) on hospital discharge and in-hospital mortality was evaluated. Of 270 SAH patients, 129 (48%) developed an infection. Depending on the underlying criteria, the incidence of sepsis and septic shock ranged between 21–46% and 9–39%. In multivariate logistic regression, the Sepsis-1 criteria were not associated with the outcome. The Sepsis-3 criteria were not associated with the functional outcome, but in shock with mortality. Alternative sepsis criteria were associated with mortality for sepsis and in shock with mortality and the functional outcome. While Sepsis-1 criteria were irrelevant for the outcome in SAH patients, septic shock, according to the Sepsis-3 criteria, adversely impacted survival. This impact was higher for the modified Sepsis-3 criteria, accounting for SAH-specific treatment. Modified Sepsis-3 and alternative sepsis criteria diagnosed septic conditions of a higher relevance for outcomes in patients with an SAH.

Development and Validation of a Model to Predict the Contract Service of Family Doctor: A National Survey in China

Background

Previous studies have reported a relatively low utilization of family doctor contract services (FDCS) in China, while the associated factors are unknown. The current study aimed to explore the factors associated with the utilization of FDCS, and then developed and validated a predictive model based on these identified factors.

Methods

We conducted a nationwide cross-sectional study using an online questionnaire, from March 2019 to April of 2019. Routinely collected variables in daily practice by family doctors were used to develop a derivation model to determine the factors associated with FDCS utilization, and then the external performance of the model was tested.

Results

A total of 115,717 and 49,593 participants were included in the development and validation datasets, respectively. Nearly 6.8% of the participants who signed a contract with FDCS received healthcare services from family doctors in China. Factors associated with the utilization of FDCS included age, male sex, self-reported household income, education attainment, insurance status, self-reported health status, smoking, drinking, self-reported physical activity status, chronic disease, walking distance from the nearest community center, and illness in the last 2 weeks, with an area under the receiver operating characteristic curve (AUC) of 0.660 [95% confidence interval (CI), 0.653–0.667] and good calibration. Application of this nomogram in the validation dataset also showed acceptable diagnostic value with an AUC of 0.659 (95% CI, 0.649–0.669) and good calibration.

Conclusion

Twelve easily obtainable factors in daily practice of family doctors were used to develop a model to predict the utilization of FDCS, with a moderate performance.

The incidence, risk factors and outcomes of acute kidney injury in critically ill patients undergoing emergency surgery: a prospective observational study

BACKGROUND

Without sufficient evidence in postoperative acute kidney injury (AKI) in critically ill patients undergoing emergency surgery, it is meaningful to explore the incidence, risk factors, and prognosis of postoperative AKI.

METHODS

A prospective observational study was conducted in the general intensive care units (ICUs) from January 2014 to March 2018. Variables about preoperation, intraoperation and postoperation were collected. AKI was diagnosed using the Kidney Disease: Improving Global Outcomes criteria.

RESULTS

Among 383 critically ill patients undergoing emergency surgery, 151 (39.4%) patients developed postoperative AKI. Postoperative reoperation, postoperative Acute Physiology and Chronic Health Evaluation (APACHE II) score, and postoperative serum lactic acid (LAC) were independent risk factors for postoperative AKI, with the adjusted odds ratio (ORadj) of 1.854 (95% confidence interval [CI], 1.091-3.152), 1.059 (95%CI, 1.018-1.102), and 1.239 (95%CI, 1.047-1.467), respectively. Compared with the non-AKI group, duration of mechanical ventilation, renal replacement therapy, ICU and hospital mortality, ICU and hospital length of stay, total ICU and hospital costs were higher in the AKI group.

CONCLUSIONS

Postoperative reoperation, postoperative APACHE II score, and postoperative LAC were independent risk factors of postoperative AKI in critically ill patients undergoing emergency surgery.

A simulation study of regression approaches for estimating risk ratios in the presence of multiple confounders

Background

Risk ratio is a popular effect measure in epidemiological research. Although previous research has suggested that logistic regression may provide biased odds ratio estimates when the number of events is small and there are multiple confounders, the performance of risk ratio estimation has yet to be examined in the presence of multiple confounders.

Methods

We conducted a simulation study to evaluate the statistical performance of three regression approaches for estimating risk ratios: (1) risk ratio interpretation of logistic regression coefficients, (2) modified Poisson regression, and (3) regression standardization using logistic regression. We simulated 270 scenarios with systematically varied sample size, the number of binary confounders, exposure proportion, risk ratio, and outcome proportion. Performance evaluation was based on convergence proportion, bias, standard error estimation, and confidence interval coverage.

Results

With a sample size of 2500 and an outcome proportion of 1%, both logistic regression and modified Poisson regression at times failed to converge, and the three approaches were comparably biased. As the outcome proportion or sample size increased, modified Poisson regression and regression standardization yielded unbiased risk ratio estimates with appropriate confidence intervals irrespective of the number of confounders. The risk ratio interpretation of logistic regression coefficients, by contrast, became substantially biased as the outcome proportion increased.

Conclusions

Regression approaches for estimating risk ratios should be cautiously used when the number of events is small. With an adequate number of events, risk ratios are validly estimated by modified Poisson regression and regression standardization, irrespective of the number of confounders.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12982-021-00107-2.

A nomogram incorporating functional and tubular damage biomarkers to predict the risk of acute kidney injury for septic patients

Background

Combining tubular damage and functional biomarkers may improve prediction precision of acute kidney injury (AKI). Serum cystatin C (sCysC) represents functional damage of kidney, while urinary N-acetyl-β-D-glucosaminidase (uNAG) is considered as a tubular damage biomarker. So far, there is no nomogram containing this combination to predict AKI in septic cohort. We aimed to compare the performance of AKI prediction models with or without incorporating these two biomarkers and develop an effective nomogram for septic patients in intensive care unit (ICU).

Methods

This was a prospective study conducted in the mixed medical-surgical ICU of a tertiary care hospital. Adults with sepsis were enrolled. The patients were divided into development and validation cohorts in chronological order of ICU admission. A logistic regression model for AKI prediction was first constructed in the development cohort. The contribution of the biomarkers (sCysC, uNAG) to this model for AKI prediction was assessed with the area under the receiver operator characteristic curve (AUC), continuous net reclassification index (cNRI), and incremental discrimination improvement (IDI). Then nomogram was established based on the model with the best performance. This nomogram was validated in the validation cohort in terms of discrimination and calibration. The decision curve analysis (DCA) was performed to evaluate the nomogram’s clinical utility.

Results

Of 358 enrolled patients, 232 were in the development cohort (69 AKI), while 126 in the validation cohort (52 AKI). The first clinical model included the APACHE II score, serum creatinine, and vasopressor used at ICU admission. Adding sCysC and uNAG to this model improved the AUC to 0.831. Furthermore, incorporating them significantly improved risk reclassification over the predictive model alone, with cNRI (0.575) and IDI (0.085). A nomogram was then established based on the new model including sCysC and uNAG. Application of this nomogram in the validation cohort yielded fair discrimination with an AUC of 0.784 and good calibration. The DCA revealed good clinical utility of this nomogram.

Conclusions

A nomogram that incorporates functional marker (sCysC) and tubular damage marker (uNAG), together with routine clinical factors may be a useful prognostic tool for individualized prediction of AKI in septic patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12882-021-02388-w.

Outcomes of critically ill end-stage kidney disease patients who underwent major surgery

Purpose

End-stage kidney disease (ESKD) is a major worldwide health problem. Patients with ESKD are thought to have a significant risk for development of complications following an operation. However, the study of ESKD and its outcomes following major operations remains rare, particularly in critical illness. Therefore, this study aimed to demonstrate how the outcomes of ESKD patients were affected when they underwent a major operation and were admitted to the intensive care unit (ICU), compared with non-ESKD patients.

Methods

A retrospective matched case cohort study was conducted in 122 critically ill surgical patients who underwent a major operation and were admitted to the ICU, during 2013 and 2016. Sixty-one ESKD patients who required long-term dialysis were enrolled and compared with 61 matched non-ESKD patients. The matching criteria were the same age interval (±5 years), gender, and type of operation. The ICU mortality was compared to the primary outcome of the study.

Results

Patients’ baseline characteristics between ESKD and non-ESKD were similar to a priori matching criteria and other demographics, except for pre-existing diabetes mellitus and hypertension, which were found significantly more in ESKD (p = 0.03 and 0.04, respectively). For operations, ESKD showed a higher grade of the American Society of Anesthesiologist (ASA) physical status (p < 0.001), but there were no differences for emergency surgery (p = 0.71) and duration of operation (p = 0.34). At ICU admission, the severity of illness measured by the Sequential Organ Failure Assessment (SOFA) score was greater in ESKD (8.9 ± 2.6 vs 5.6 ± 2.5; p < 0.001). However, after eliminating renal domain, SOFA non-renal score was equivalent (5.7 ± 2.2 vs 5.2 ± 2.3, p = 0.16). The ICU mortality was significantly higher in critically-ill surgical patients with ESKD than non-ESKD (23% vs 5%, p=0.007), along with hospital mortality rates (34% vs 10%, p = 0.002). The multivariable logistic regression analyses adjusted for age and SOFA non-renal score demonstrated that ESKD had a significant association with ICU and hospital mortality (adjOR = 5.59; 95%CI [1.49–20.88], p = 0.01 and adjOR = 4.55; 95%CI[1.67–12.44], p = 0.003, respectively).

Conclusion

Patients who underwent a major operation and needed intensive care admission with pre-existing ESKD requiring long-term dialysis were associated with greater mortality than patients without ESKD. More careful assessment before, during, and after major surgical procedures should be performed in this group of patients to improve post-operative outcomes.

Adaptive sample size determination for the development of clinical prediction models

BACKGROUND

We suggest an adaptive sample size calculation method for developing clinical prediction models, in which model performance is monitored sequentially as new data comes in.

METHODS

We illustrate the approach using data for the diagnosis of ovarian cancer (n = 5914, 33% event fraction) and obstructive coronary artery disease (CAD; n = 4888, 44% event fraction). We used logistic regression to develop a prediction model consisting only of a priori selected predictors and assumed linear relations for continuous predictors. We mimicked prospective patient recruitment by developing the model on 100 randomly selected patients, and we used bootstrapping to internally validate the model. We sequentially added 50 random new patients until we reached a sample size of 3000 and re-estimated model performance at each step. We examined the required sample size for satisfying the following stopping rule: obtaining a calibration slope ≥ 0.9 and optimism in the c-statistic (or AUC) < = 0.02 at two consecutive sample sizes. This procedure was repeated 500 times. We also investigated the impact of alternative modeling strategies: modeling nonlinear relations for continuous predictors and correcting for bias on the model estimates (Firth's correction).

RESULTS

Better discrimination was achieved in the ovarian cancer data (c-statistic 0.9 with 7 predictors) than in the CAD data (c-statistic 0.7 with 11 predictors). Adequate calibration and limited optimism in discrimination was achieved after a median of 450 patients (interquartile range 450-500) for the ovarian cancer data (22 events per parameter (EPP), 20-24) and 850 patients (750-900) for the CAD data (33 EPP, 30-35). A stricter criterion, requiring AUC optimism < = 0.01, was met with a median of 500 (23 EPP) and 1500 (59 EPP) patients, respectively. These sample sizes were much higher than the well-known 10 EPP rule of thumb and slightly higher than a recently published fixed sample size calculation method by Riley et al. Higher sample sizes were required when nonlinear relationships were modeled, and lower sample sizes when Firth's correction was used.

CONCLUSIONS

Adaptive sample size determination can be a useful supplement to fixed a priori sample size calculations, because it allows to tailor the sample size to the specific prediction modeling context in a dynamic fashion.

Admission Levels of Interleukin 10 and Amyloid β 1-40 Improve the Outcome Prediction Performance of the Helsinki Computed Tomography Score in Traumatic Brain Injury

Background: Blood biomarkers may enhance outcome prediction performance of head computed tomography scores in traumatic brain injury (TBI).

Objective: To investigate whether admission levels of eight different protein biomarkers can improve the outcome prediction performance of the Helsinki computed tomography score (HCTS) without clinical covariates in TBI.

Materials and methods: Eighty-two patients with computed tomography positive TBIs were included in this study. Plasma levels of β-amyloid isoforms 1–40 (Aβ40) and 1–42 (Aβ42), glial fibrillary acidic protein, heart fatty acid-binding protein, interleukin 10 (IL-10), neurofilament light, S100 calcium-binding protein B, and total tau were measured within 24 h from admission. The patients were divided into favorable (Glasgow Outcome Scale—Extended 5–8, n = 49) and unfavorable (Glasgow Outcome Scale—Extended 1–4, n = 33) groups. The outcome was assessed 6–12 months after injury. An optimal predictive panel was investigated with the sensitivity set at 90–100%.

Results: The HCTS alone yielded a sensitivity of 97.0% (95% CI: 90.9–100) and specificity of 22.4% (95% CI: 10.2–32.7) and partial area under the curve of the receiver operating characteristic of 2.5% (95% CI: 1.1–4.7), in discriminating patients with favorable and unfavorable outcomes. The threshold to detect a patient with unfavorable outcome was an HCTS > 1. The three best individually performing biomarkers in outcome prediction were Aβ40, Aβ42, and neurofilament light. The optimal panel included IL-10, Aβ40, and the HCTS reaching a partial area under the curve of the receiver operating characteristic of 3.4% (95% CI: 1.7–6.2) with a sensitivity of 90.9% (95% CI: 81.8–100) and specificity of 59.2% (95% CI: 40.8–69.4).

Conclusion: Admission plasma levels of IL-10 and Aβ40 significantly improve the prognostication ability of the HCTS after TBI.

Patency Rates After Successful Arteriovenous Fistula Thrombectomy: Relevance of the Flow/d-Dimer Ratio in the Decision-Making

OBJECTIVES

Surgical thrombectomy for acute arteriovenous fistula (AVF) thrombosis is one of the primary salvage intervention. The independent risk factors affecting the patency of AVF after a successful thrombectomy are yet unknown. Here, the author aimed to report the results of surgically corrected AVFs and the independent risk factors which may cause early failure following the surgical salvage.

METHODS

The study cohort comprised 24 patients who had acute AVF thrombosis and underwent successful surgical thrombectomy in the first 24 to 48 hours between January 2016 and April 2020 in our center. The study group was divided into patients with recurrent AVF thrombosis (n = 11, 45.8%) and without recurrent AVF thrombosis (n = 13, 54.1%) following surgical thrombectomy with a follow-up of 22.4 ± 6.8 months. Postthrombectomy primary and secondary patency of AVF were also evaluated.

RESULTS

The mean age of the cohort was 58.1 ± 15.2 years. A simple thrombectomy was performed for all cases. Only 2 cases have required a revision at the anastomosis due to severe intimal hyperplasia. Postthrombectomy primary patency rate was 45.5% for 18 months. Receiver operating characteristic analysis was performed with a resulting area under the curve value of 0.81 (95% CI: 0.35-0.94, P = .006) for flow (mL)/d-dimer (ng/mL) <0.63 in predicting recurrent AVF thrombosis following surgical thrombectomy.

CONCLUSIONS

Flow (mL)/d-dimer (ng/mL) <0.63 was independent predictor of recurrent thrombosis (RT) of a surgically salvaged AVF. The patients at risk for RT or who may benefit from further intervention should be identified with predictive parameters.

Preoperative prediction of Bleeding Independently associated with Mortality after noncardiac Surgery (BIMS): an international prospective cohort study

BACKGROUND

Diagnostic criteria for Bleeding Independently associated with Mortality after noncardiac Surgery (BIMS) have been defined as bleeding that leads to a postoperative haemoglobin <70 g L^-1, leads to blood transfusion, or is judged to be the direct cause of death. Preoperative prediction guides for BIMS can facilitate informed consent and planning of perioperative care.

METHODS

In a prospective cohort study of 16 079 participants aged ≥45 yr having inpatient noncardiac surgery at 12 academic hospitals in eight countries between 2007 and 2011, 17.3% (2782) experienced BIMS. An electronic risk calculator for BIMS was developed and internally validated by logistic regression with bootstrapping, and further simplified to a risk index. Decision curve analysis assessed the potential utility of each prediction guide compared with a strategy of identifying risk of BIMS based on preoperative haemoglobin <120 g L^-1.

RESULTS

With information about the type of surgery, preoperative haemoglobin, age, sex, functional status, kidney function, history of high-risk coronary artery disease, and active cancer, the risk calculator accurately predicted BIMS (bias-corrected C-statistic, 0.84; 95% confidence interval, 0.837-0.852). A simplified index based on preoperative haemoglobin <120 g L^-1, open surgery, and high-risk surgery also predicted BIMS, but less accurately (C-statistic, 0.787; 95% confidence interval, 0.779-0.796). Both prediction guides could improve decision making compared with knowledge of haemoglobin <120 g L^-1 alone.

CONCLUSIONS

BIMS, defined as bleeding that leads to a postoperative haemoglobin <70 g L^-1, leads to blood transfusion, or that is judged to be the direct cause of death, can be predicted by a simple risk index before surgery.

CLINICAL TRIAL REGISTRATION

NCT00512109.

Geographical differences in cancer treatment and survival for patients with oesophageal and gastro-oesophageal junctional cancers

BACKGROUND

Only around one-quarter of patients with cancer of the oesophagus and the gastro-oesophageal junction (GOJ) undergo surgical resection. This population-based study investigated the rates of treatment with curative intent and resection, and their association with survival.

METHODS

Patients diagnosed with oesophageal and GOJ cancer between 2006 and 2015 in Sweden were identified from the National Register for Oesophageal and Gastric Cancer (NREV). The NREV was cross-linked with several national registries to obtain information on additional exposures. The annual proportion of patients undergoing treatment with curative intent and surgical resection in each county was calculated, and the counties divided into groups with low, intermediate and high rates. Treatment with curative intent was defined as definitive chemoradiation therapy or surgery, with or without neoadjuvant oncological treatment. Overall survival was analysed using a multilevel model based on county of residence at the time of diagnosis.

RESULTS

Some 5959 patients were included, of whom 1503 (25·2 per cent) underwent surgery. Median overall survival after diagnosis was 7·7, 8·8 and 11·1 months respectively in counties with low, intermediate and high rates of treatment with curative intent. Corresponding survival times for the surgical resection groups were 7·4, 9·3 and 11·0 months. In the multivariable analysis, a higher rate of treatment with curative intent (time ratio 1·17, 95 per cent c.i. 1·05 to 1·30; P < 0·001) and a higher resection rate (time ratio 1·24, 1·12 to 1·37; P < 0·001) were associated with improved survival after adjustment for relevant confounders.

CONCLUSION

Patients diagnosed in counties with higher rates of treatment with curative intent and higher rates of surgery had better survival.

Nomogram for contrast-induced acute kidney injury in patients with chronic kidney disease undergoing coronary angiography in China: a cohort study

OBJECTIVES

To establish a nomogram for contrast-induced acute kidney injury (CI-AKI) risk assessment among patients with chronic kidney disease (CKD) undergoing coronary angiography (CAG) or percutaneous coronary intervention (PCI).

DESIGN

Prospective observational cohort study.

SETTING

Southern China.

INTERVENTIONS

None.

PARTICIPANTS

643 consecutive patients with CKD (defined as estimated glomerular filtration rate calculated by Modification of Diet in Renal Disease formula <60 mL/min/1.73 mm²) were enrolled.

OUTCOME MEASURES

The end point was CI-AKI defined as serum creatinine elevation ≥0.5 mg/dL or 25% from baseline within the first 48-72 hours following contrast exposure.Predictors of CI-AKI were selected by multivariable logistic regression and stepwise approach. A nomogram based on these predictors was constructed and compared with the classic Mehran Score. For validation, a bootstrap method (1000 times) was performed.

RESULTS

The nomogram including age, weight, heart rate, hypotension, PCI and β-blocker demonstrated a better predictive value than the classic Mehran Score (area under the curve: 0.78 vs 0.71, p=0.024), as well as a well-fitted calibration curve (χ²=12.146, p=0.145). Validation through the bootstrap method (1000 times) also indicated a good discriminative power (adjusted C-statistic: 0.76).

CONCLUSIONS

With fewer predictors and higher discriminative power, the present nomogram may be a simple and reliable tool to identify patients with CKD at risk of CI-AKI, whereas further external validations are needed.

In situ cell cycle analysis in giant cell tumor of bone reveals patients with elevated risk of reduced progression-free survival

OBJECTIVE

Giant cell tumor of bone (GCTB) is a frequently recurring locally aggressive osteolytic lesion, where pathological osteoclastogenesis and bone destruction are driven by neoplastic stromal cells. Here, we studied if cell cycle fractions within the mononuclear cell compartment of GCTB can predict its progression-free survival (PFS).

METHODS

154 cases (100 primaries and 54 recurrent) from 139 patients of 40 progression events, was studied using tissue microarrays. Ploidy and in situ cell cycle progression related proteins including Ki67 and those linked with replication licensing (mcm2), G1-phase (cyclin D1, Cdk4), and S-G2-M-phase (cyclin A; Cdk2) fractions; cell cycle control (p21^waf1) and repression (geminin), were tested. The Prentice-Williams-Peterson (PWP) gap-time models with the Akaike information criterion (AIC) were used for PFS analysis.

RESULTS

Cluster analysis showed good correlation between functionally related marker positive cell fractions indicating no major cell cycle arrested cell populations in GCTB. Increasing hazard of progression was statistically associated with the elevated post-G1/S-phase cell fractions. Univariate analysis revealed significant negative association of poly-/aneuploidy (p < 0.0001), and elevated cyclin A (p < 0.001), geminin (p = 0.015), mcm2 (p = 0.016), cyclin D1 (p = 0.022) and Ki67 (B56: p = 0.0543; and Mib1: p = 0.0564 -strong trend) positive cell fractions with PFS. The highest-ranked multivariate interaction model (AIC = 269.5) also included ploidy (HR 5.68, 95%CI: 2.62-12.31, p < 0.0001), mcm2 (p = 0.609), cyclin D1 (HR 1.89, 95%CI: 0.88-4.09, p = 0.105) and cyclin A (p < 0.0001). The first and second best prognostic models without interaction (AIC = 271.6) and the sensitivity analysis (AIC = 265.7) further confirmed the prognostic relevance of combining these markers.

CONCLUSION

Ploidy and elevated replication licensing (mcm2), G1-phase (cyclin D1) and post-G1 phase (cyclin A) marker positive cell fractions, indicating enhanced cell cycle progression, can assist in identifying GCTB patients with increased risk for a reduced PFS.

Methodological standards for the development and evaluation of clinical prediction rules: a review of the literature

Clinical prediction rules (CPRs) that predict the absolute risk of a clinical condition or future outcome for individual patients are abundant in the medical literature; however, systematic reviews have demonstrated shortcomings in the methodological quality and reporting of prediction studies. To maximise the potential and clinical usefulness of CPRs, they must be rigorously developed and validated, and their impact on clinical practice and patient outcomes must be evaluated. This review aims to present a comprehensive overview of the stages involved in the development, validation and evaluation of CPRs, and to describe in detail the methodological standards required at each stage, illustrated with examples where appropriate. Important features of the study design, statistical analysis, modelling strategy, data collection, performance assessment, CPR presentation and reporting are discussed, in addition to other, often overlooked aspects such as the acceptability, cost-effectiveness and longer-term implementation of CPRs, and their comparison with clinical judgement. Although the development and evaluation of a robust, clinically useful CPR is anything but straightforward, adherence to the plethora of methodological standards, recommendations and frameworks at each stage will assist in the development of a rigorous CPR that has the potential to contribute usefully to clinical practice and decision-making and have a positive impact on patient care.

Assessment of right heart functions in the patients with arteriovenous fistula for hemodialysis access: Right ventricular free wall strain and tricuspid regurgitation jet velocity as the predictors of right heart failure

Objectives

Previous studies have mostly focused on the left-sided cardiovascular changes, but right-sided cardiac changes and predictive factors have not been examined in advance following arteriovenous fistula (AVF) creation. We aimed to identify new parameters which contribute to the prediction of right heart failure (RHF) after AVF creation.

Methods

The study cohort comprised 81 patients who underwent AVF creation between January 2014 and April 2019 in our center. The study cohort was divided into the patients with RHF ( n = 15, 18.5%) and without RHF ( n = 66, 72.5%) following AVF creation.

Results

Mean age of cohort was 49.9 ± 14.7 years (range 23–66) and 39 (48.1%) were men. Approximately 74.07% (60 patients) were in New York Heart Association Class II and III profile preoperatively. Independent predictors for RHF following AVF were right ventricle longitudinal strain (RVLS) free wall <-19% [odds ratio (OR) 2.31, 95% CI 1.02–3.22], and tricuspid regurgitation jet velocity (TRJV) >2.5 m/s [odds ratio (OR) 5.68, 95% CI 1.21–4.38]. Receiver operating characteristic analysis was performed with a resulting area under the curve value of 0.86 (95% CI 0.55–0.89, p = 0.004) for RVLS free wall <-14.2% and 0.81 for TRJV >2.61 m/s (95% CI 0.55–0.89, p = 0.005) in predicting RHF following AVF.

Conclusions

RVLS free wall <-14.2% and TRJV >2.61 m/s were independent predictors of RHF following AVF creation. The patients at risk for having RHF following AVF creation or who may benefit from AVF should be identified with predictive parameters and prospective clinical studies.

Predicting individualized mortality probabilities for patients with squamous cell carcinoma of the maxilla: Novel models with clinical and histopathological predictors

BACKGROUND

The aim of this article was to develop prediction models that calculate postoperative 2- and 5-year mortality probabilities of patients with squamous cell carcinoma of the maxilla (MSCC).

METHODS

Data were collected from the medical records of patients who had been operated between 2000 and 2015 for MSCC. Potential clinical and histopathological predictors were identified. Confounding-(un)adjusted multivariate Cox and logistic regression models were computed with stepwise backward selection. Internal validation was performed to assess calibration and discriminatory ability.

RESULTS

Ninety-five patients with MSCC were included. Two-year follow-up was complete, and 85 patients had 5-year follow-up. Age, neck treatment, surgical margins, bone invasion, spindle growth, and vasoinvasive growth were associated with mortality. Models were adjusted for confounding with Charlson's comorbidities index. C-indexes were .841 and .770 respectively, and .838 and .749 after bootstrapping.

CONCLUSION

The MSCC-specific mortality probability can be calculated with new prediction models.

Predicting Severe Cartilage Damage in the Hip: A Model Using Patient-Specific Data From 2,396 Hip Arthroscopies

PURPOSE

To determine patient-specific factors that can be used to predict the presence of severe articular cartilage damage in the hip in patients without osteoarthritis.

METHODS

The prevalence of severe (Outerbridge grade III or IV) cartilage damage to the acetabulum and femoral head was prospectively recorded at hip arthroscopy. Patients who underwent primary hip arthroscopic surgery between 2006 and 2016 performed by a single surgeon were included. Patients were excluded if they underwent previous hip surgery, had poor-quality radiographs, were younger than 16 years at the time of surgery, or had a minimal joint space of 2 mm or less. The relation between severe cartilage damage and preoperative patient characteristics was examined using multivariable logistic regression analysis with restricted cubic splines.

RESULTS

Of the 2,396 hips presenting for hip arthroscopy, 995 (41%) had severe cartilage damage to the acetabulum and 257 (11%) had severe cartilage damage to the femoral head. Older age was a significant risk factor for severe cartilage damage both to the acetabulum (χ² = 69.5, P < .001) and to the femoral head (χ² = 53.9, P < .001). An age of 45 years was associated with a 1.96 (95% confidence interval, 1.54-2.49) increase in the odds of severe acetabular cartilage damage and a 3.94 (95% confidence interval, 2.61-5.94) increase in the odds of severe femoral head cartilage damage relative to an age of 20 years. Male sex was associated with severe cartilage damage to the acetabulum (χ² = 66.7, P < .001), and a lower center-edge angle was a significant risk factor for severe cartilage damage to the femoral head (χ² = 78.5, P < .001). Predictive nomograms were established for severe cartilage lesions.

CONCLUSIONS

The primary risk factors for severe hip cartilage damage were older age for both the femoral head and acetabulum; a lower center-edge angle and larger Tönnis angle for the femoral head; and male sex, body mass index, alpha angle, and joint space for the acetabulum. The likelihood of cartilage damage to the hip can be estimated clinically using a prediction nomogram.

LEVEL OF EVIDENCE

Level III, cross-sectional study.

I thought I saw a pussy cat: Portrayal of wild cats in friendly interactions with humans distorts perceptions and encourages interactions with wild cat species

Most people lack the opportunity to see non-domesticated animals in the wild. Consequently, people’s perception of wild animals is based on what they see on (social) media. The way in which (social) media portrays non-domesticated animals determines our perception of and behaviour to these animals. People like to interact with animals, which is why venues which offer the opportunity to interact with non-domesticated animals are popular wildlife tourist attractions (WTAs). However, these WTAs more often than not profit at the expense of animal welfare, conservation and human safety. Participation in such WTAs should therefore be discouraged. Through (social) media we are regularly exposed to images of non-domesticated animals in close interactions with humans. Exposure to such images seems to blur the line between what is a friendly domesticated animal and what is a potentially dangerous wild animal. Such images may also increase our desire to engage in interactions with non-domesticated animals ourselves and reduce moral concerns about the use of non-domesticated animals for such interactions, thereby promoting WTAs in which tourists can interact with non-domesticated animals. Wild cat species are commonly used in the wildlife tourism industry to interact with tourists. In this study, we determine whether portrayal of wild cat species in interactions with humans promotes WTAs with wild cats. We presented respondents with an image of a wild cat species (lion, cheetah, caracal) in a control setting, walked by a human (WTA), petted by a human (WTA) or in the wild and asked them to answer a fixed set of questions. We found that portraying wild cat species in interactions with humans reduced the fear of wild cats, encouraged people to regard WTAs with wild cats as acceptable and stimulated them to participate in such activities themselves.

An Integrated Investigation of Spatiotemporal Habitat Quality Dynamics and Driving Forces in the Upper Basin of Miyun Reservoir, North China

Understanding changes in habitat quality and the driving forces of these changes at landscape scales is a critical part of effective ecosystem management. The present study investigated spatiotemporal habitat quality dynamics and related driving forces from 2005 to 2015 in the upper basin of Miyun Reservoir in North China by constructing an effective framework integrated InVEST and binary logistic regression models. This framework expanded the driving force analysis into an assessment of changes in habitat quality and intuitively verified the effectiveness of relevant environmental policies. The proposed method of combining the equidistant random sampling method and the method of introducing spatial lag variables in logistic regression equation can effectively solve spatial autocorrelation with a large enough number of sampling points. Overall, habitat quality improved during the study period. Spatially, a concentrated loss of habitat occurred in the southeastern part of the basin between the reservoir and mountainous areas, while other areas gradually recovered. Driving force analysis showed that lower elevation mountain land, gentle slopes, locations near rural land or roads, larger areas of grain cultivation, and areas with little population change had a higher likelihood of having changed in habitat quality in the upper basin of Miyun Reservoir. These results suggested that the present policy of protecting the ecosystem had a positive effect on improving habitat quality. In the future, the human activity management related to habitat quality needs to be strengthened. The present study would provide a reference for land use policy formulation and biodiversity conservation.

A prospective cohort study of early-pregnancy risk factors for gestational diabetes in polycystic ovarian syndrome

BACKGROUND

Polycystic ovarian syndrome (PCOS) is a strong risk factor for gestational diabetes (GDM). However, the association between features of PCOS during early pregnancy and the risk of GDM is not clearly characterized. In this prospective cohort study, we seek to identify early-pregnancy risk factors for GDM in PCOS women.

METHODS

Between 2011 and 2013, 248 women with PCOS were followed from their first prenatal visit to delivery. Multiple early-pregnancy metabolic factors were evaluated for their association with the risk of GDM.

RESULTS

Among 248 subjects, 75 (30.2%) developed GDM. Single factor analysis identified a number of metabolic risk factors for GDM, including higher body mass index, fasting plasma glucose (FPG) and insulin resistance; abnormal cholesterol; elevated blood pressure and free androgen index; lower level of sex-hormone binding globulin (SHBG); and less gestational weight gain. Multivariate analysis showed that FPG, non-high-density lipoprotein-cholesterol and SHBG are independent predictive factors for GDM.

CONCLUSIONS

Our study established strong association of multiple early-pregnancy risk factors with development of GDM in PCOS women. These risk factors are predominantly related to the regulation of glucose, lipid, and androgen metabolism. Among these factors, FPG, non-high-density lipoprotein-cholesterol, and SHBG, predict incident GDM.

The Incidence, Risk Factors and Outcomes of Postoperative Acute Kidney Injury in Neurosurgical Critically Ill Patients

We investigated the incidence, perioperative risk factors, and outcomes of postoperative acute kidney injury (AKI) in neurosurgical critically ill patients. A prospective multicenter cohort study was conducted, enrolling adult patients who underwent neurosurgical procedure and admitted to the neurosurgical intensive care units (ICU). Postoperative AKI was diagnosed within 7 days after surgery based on the Kidney Disease Improving Global Outcomes criteria. Of 624 enrolled patients, postoperative AKI occurred in 84 patients. AKI was associated with increased rates of ICU and in-hospital mortality, postoperative renal replacement therapy, postoperative tracheotomy, and postoperative tracheal reintubation. Patients who developed AKI had higher total ICU costs, prolonged length of hospital and ICU stay, and longer duration of postoperative mechanical ventilation. Multivariate analysis identified postoperative reoperation (adjusted odds ratio [OR] 5.70 [95% CI, 1.61–20.14]), postoperative concentration of serum cystatin C (adjusted OR 4.53 [95% CI, 1.98–10.39]), use of mannitol during operation (adjusted OR 1.97 [95% CI, 1.13–3.43]), postoperative APACHE II score (adjusted OR 1.11 [95% CI, 1.06–1.16]), and intraoperative estimated blood loss (adjusted OR 1.04 [95% CI, 1.00–1.08]) as independent risk factors for postoperative AKI. Postoperative AKI in neurosurgical critically ill cohort is prevalent and associated with adverse in-hospital outcomes.

Predictors of Short- and Long-Term Attrition From the Parents as Agents of Change Randomized Controlled Trial for Managing Pediatric Obesity

INTRODUCTION

Attrition in pediatric weight management is a substantial problem. This study examined factors associated with short- and long-term attrition from a lifestyle and behavioral intervention for parents of children with overweight or obesity.

METHOD

Fifty-two families with children ages 6 to 12 years old and body mass index at or above the 85th percentile participated in a randomized controlled trial focused on parents, comparing parent-based cognitive behavioral therapy with parent-based psychoeducation for pediatric weight management. We examined program attrition using two clinical phases of the intervention: short-term and long-term attrition, modeled using the general linear model. Predictors included intervention type, child/parent weight status, sociodemographic factors, and health of the family system.

RESULTS

Higher self-assessed health of the family system was associated with lower short-term attrition; higher percentage of intervention sessions attended by parents was associated with lower long-term attrition.

DISCUSSION

Different variables were significant in our short- and long-term models. Attrition might best be conceptualized based on short- and long-term phases of clinical, parent-based interventions for pediatric weight management.

New-Onset Atrial Fibrillation in the Critically Ill

OBJECTIVE

To determine the association of new-onset atrial fibrillation with outcomes, including ICU length of stay and survival.

DESIGN

Retrospective cohort of ICU admissions. We found atrial fibrillation using automated detection (≥ 90 s in 30 min) and classed as new-onset if there was no prior diagnosis of atrial fibrillation. We identified determinants of new-onset atrial fibrillation and, using propensity matching, characterized its impact on outcomes.

SETTING

Tertiary care academic center.

PATIENTS

A total of 8,356 consecutive adult admissions to either the medical or surgical/trauma/burn ICU with available continuous electrocardiogram data.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

From 74 patient-years of every 15-minute observations, we detected atrial fibrillation in 1,610 admissions (19%), with median burden less than 2%. Most atrial fibrillation was paroxysmal; less than 2% of admissions were always in atrial fibrillation. New-onset atrial fibrillation was subclinical or went undocumented in 626, or 8% of all ICU admissions. Advanced age, acute respiratory failure, and sepsis were the strongest predictors of new-onset atrial fibrillation. In propensity-adjusted regression analyses, clinical new-onset atrial fibrillation was associated with increased hospital mortality (odds ratio, 1.63; 95% CI, 1.01-2.63) and longer length of stay (2.25 d; CI, 0.58-3.92). New-onset atrial fibrillation was not associated with survival after hospital discharge (hazard ratio, 0.99; 95% CI, 0.76-1.28 and hazard ratio, 1.11; 95% CI, 0.67-1.83, respectively, for subclinical and clinical new-onset atrial fibrillation).

CONCLUSIONS

Automated analysis of continuous electrocardiogram heart rate dynamics detects new-onset atrial fibrillation in many ICU patients. Though often transient and frequently unrecognized, new-onset atrial fibrillation is associated with poor hospital outcomes.

Postoperative Bleeding After Change in Heparin Supplier: A Cardiothoracic Center Experience

OBJECTIVE

Unfractionated heparin is a mixture of glycosaminoglycans with different pharmacologic and pharmacokinetic properties. The literature suggests that blood loss after cardiac surgery is related to both elevated postoperative heparin concentrations and the potency of different heparin brands.

DESIGN

An audit of the observed increase in the incidence of cardiac surgery-related bleeding after change in heparin supplier. Patient characteristics were compared between groups before and after a change in heparin brands.

SETTING

Tertiary cardiothoracic center.

PARTICIPANTS

All patients undergoing cardiac surgery between August 1, 2011, and April 30, 2012.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Two hundred eighty patients underwent surgery before a change in heparin brands and 216 after a change. Their preoperative and intraoperative characteristics were similar. Postoperative chest tube drainages and blood transfusions were significantly greater after the change in heparin brands (postoperative chest drainage 476.8 ± 393.1 v 344.8 ± 323.2 mL/6 h and 1,062.2 ± 738.8 v 841.8 ± 567.4 mL/24 h, respectively; both p < 0.001) despite the administration of larger amounts of protamine, fresh frozen plasma/platelet transfusions, and cryoprecipitate. Heparin recirculation within 24 hours of bypass was noted in about 70% of the samples tested using either anti-factor X activity or the thromboelastography ratio between nonheparinase R and heparinase-modified R and was not associated with the heparin brand. The likelihood ratio chi-square test for nested models identified an added predictive value of the heparin brand when included as a predictor of bleeding (chest drainage >800 mL/6 h) in a model comprising recirculation, assessed using either an elevated anti-factor X activity or ratio between nonheparinase R and heparinase-modified R.

CONCLUSION

It is likely that the observed increase in postoperative bleeding was related to the pharmacologic properties of the new heparin brand rather than a higher incidence of heparin recirculation.

No rationale for 1 variable per 10 events criterion for binary logistic regression analysis

BACKGROUND

Ten events per variable (EPV) is a widely advocated minimal criterion for sample size considerations in logistic regression analysis. Of three previous simulation studies that examined this minimal EPV criterion only one supports the use of a minimum of 10 EPV. In this paper, we examine the reasons for substantial differences between these extensive simulation studies.

METHODS

The current study uses Monte Carlo simulations to evaluate small sample bias, coverage of confidence intervals and mean square error of logit coefficients. Logistic regression models fitted by maximum likelihood and a modified estimation procedure, known as Firth's correction, are compared.

RESULTS

The results show that besides EPV, the problems associated with low EPV depend on other factors such as the total sample size. It is also demonstrated that simulation results can be dominated by even a few simulated data sets for which the prediction of the outcome by the covariates is perfect ('separation'). We reveal that different approaches for identifying and handling separation leads to substantially different simulation results. We further show that Firth's correction can be used to improve the accuracy of regression coefficients and alleviate the problems associated with separation.

CONCLUSIONS

The current evidence supporting EPV rules for binary logistic regression is weak. Given our findings, there is an urgent need for new research to provide guidance for supporting sample size considerations for binary logistic regression analysis.

Propensity score model overfitting led to inflated variance of estimated odds ratios

OBJECTIVE

Simulation studies suggest that the ratio of the number of events to the number of estimated parameters in a logistic regression model should be not less than 10 or 20 to 1 to achieve reliable effect estimates. Applications of propensity score approaches for confounding control in practice, however, do often not consider these recommendations.

STUDY DESIGN AND SETTING

We conducted extensive Monte Carlo and plasmode simulation studies to investigate the impact of propensity score model overfitting on the performance in estimating conditional and marginal odds ratios using different established propensity score inference approaches. We assessed estimate accuracy and precision as well as associated type I error and type II error rates in testing the null hypothesis of no exposure effect.

RESULTS

For all inference approaches considered, our simulation study revealed considerably inflated standard errors of effect estimates when using overfitted propensity score models. Overfitting did not considerably affect type I error rates for most inference approaches. However, because of residual confounding, estimation performance and type I error probabilities were unsatisfactory when using propensity score quintile adjustment.

CONCLUSION

Overfitting of propensity score models should be avoided to obtain reliable estimates of treatment or exposure effects in individual studies.

Adequate sample size for developing prediction models is not simply related to events per variable

OBJECTIVES

The choice of an adequate sample size for a Cox regression analysis is generally based on the rule of thumb derived from simulation studies of a minimum of 10 events per variable (EPV). One simulation study suggested scenarios in which the 10 EPV rule can be relaxed. The effect of a range of binary predictors with varying prevalence, reflecting clinical practice, has not yet been fully investigated.

STUDY DESIGN AND SETTING

We conducted an extended resampling study using a large general-practice data set, comprising over 2 million anonymized patient records, to examine the EPV requirements for prediction models with low-prevalence binary predictors developed using Cox regression. The performance of the models was then evaluated using an independent external validation data set. We investigated both fully specified models and models derived using variable selection.

RESULTS

Our results indicated that an EPV rule of thumb should be data driven and that EPV ≥ 20 ​ generally eliminates bias in regression coefficients when many low-prevalence predictors are included in a Cox model.

CONCLUSION

Higher EPV is needed when low-prevalence predictors are present in a model to eliminate bias in regression coefficients and improve predictive accuracy.

Post-traumatic stress disorder associated with life-threatening motor vehicle collisions in the WHO World Mental Health Surveys

BACKGROUND

Motor vehicle collisions (MVCs) are a substantial contributor to the global burden of disease and lead to subsequent post-traumatic stress disorder (PTSD). However, the relevant literature originates in only a few countries, and much remains unknown about MVC-related PTSD prevalence and predictors.

METHODS

Data come from the World Mental Health Survey Initiative, a coordinated series of community epidemiological surveys of mental disorders throughout the world. The subset of 13 surveys (5 in high income countries, 8 in middle or low income countries) with respondents reporting PTSD after life-threatening MVCs are considered here. Six classes of predictors were assessed: socio-demographics, characteristics of the MVC, childhood family adversities, MVCs, other traumatic experiences, and respondent history of prior mental disorders. Logistic regression was used to examine predictors of PTSD. Mental disorders were assessed with the fully-structured Composite International Diagnostic Interview using DSM-IV criteria.

RESULTS

Prevalence of PTSD associated with MVCs perceived to be life-threatening was 2.5 % overall and did not vary significantly across countries. PTSD was significantly associated with low respondent education, someone dying in the MVC, the respondent or someone else being seriously injured, childhood family adversities, prior MVCs (but not other traumatic experiences), and number of prior anxiety disorders. The final model was significantly predictive of PTSD, with 32 % of all PTSD occurring among the 5 % of respondents classified by the model as having highest PTSD risk.

CONCLUSION

Although PTSD is a relatively rare outcome of life-threatening MVCs, a substantial minority of PTSD cases occur among the relatively small proportion of people with highest predicted risk. This raises the question whether MVC-related PTSD could be reduced with preventive interventions targeted to high-risk survivors using models based on predictors assessed in the immediate aftermath of the MVCs.

MiR-320e is a novel prognostic biomarker in colorectal cancer

Background:

Advances in early detection and treatment have improved outcomes in patients with colorectal cancer (CRC). However, there remains a need for robust prognostic and predictive biomarkers. We conducted a systematic discovery and validation of microRNA (miRNA) biomarkers in two clinical trial cohorts of CRC patients.

Methods:

We performed an initial ‘discovery' phase using Affymetrix miRNA expression arrays to profile stage III CRC patients with and without tumour recurrence (n=50 per group) at 3-years of follow-up. All patients received adjuvant 5-fluorouracil (5-FU) plus oxaliplatin, that is, FOLFOX, treatment. During ‘validation', we analysed miRNAs using qRT–PCR in an independent cohort of 237 stage II–IV CRC patients treated with 5-FU-based chemotherapy, as well as in normal colonic mucosa from 20 healthy subjects. Association with disease recurrence, disease-free survival (DFS) and overall survival (OS) was examined using Cox proportional hazard models.

Results:

In the discovery cohort, miR-320e expression was significantly elevated in stage III colon cancers from patients with vs without recurrence (95% confidence interval (CI)=1.14–1.42; P<0.0001). These results were then independently validated in stage II and III tumours. Specifically, increased miR-320e expression was associated with poorer DFS (hazard ratio (HR)=1.65; 95% CI=1.27–2.13; P=0.0001) and OS (HR=1.78; 95% CI=1.31–2.41; P=0.0003) in stage III CRC patients.

Conclusions:

In two clinical trial cohorts, a systematic biomarker discovery and validation approach identified miR-320e to be a novel prognostic biomarker that is associated with adverse clinical outcome in stage III CRC patients treated with 5-FU-based adjuvant chemotherapy. These findings have important implications for the personalised management of CRC patients.

Statistical and data reporting guidelines for the European Journal of Cardio-Thoracic Surgery and the Interactive CardioVascular and Thoracic Surgery

As part of the peer review process for the European Journal of Cardio-Thoracic Surgery (EJCTS) and the Interactive CardioVascular and Thoracic Surgery (ICVTS), a statistician reviews any manuscript that includes a statistical analysis. To facilitate authors considering submitting a manuscript and to make it clearer about the expectations of the statistical reviewers, we present up-to-date guidelines for authors on statistical and data reporting specifically in these journals. The number of statistical methods used in the cardiothoracic literature is vast, as are the ways in which data are presented. Therefore, we narrow the scope of these guidelines to cover the most common applications submitted to the EJCTS and ICVTS, focusing in particular on those that the statistical reviewers most frequently comment on.

Proposed research plan for the derivation of a new Cardiac Risk Index

The Revised Cardiac Risk Index (RCRI) was incorporated into the American College of Cardiology/American Heart Association (ACC/AHA) recommendations for the preoperative evaluation of the cardiac patient for noncardiac surgery. The purpose of this review was to analyze studies on cardiovascular clinical risk prediction that had used the previous "standard best" model, the RCRI, as a comparator. This review aims to determine whether modification of the current risk factors or adoption of other risk factors or other risk indices would improve upon the discrimination of cardiac risk prediction when compared with the RCRI. This is necessary because recent risk prediction models have shown better discrimination for major adverse cardiac events, and the pre-eminence of the RCRI is now in question. There is now a need for a new "best standard" cardiovascular risk prediction model to supersede the RCRI. This is desirable because it would: (1) allow for a global standard of cardiovascular risk assessment; (2) provide a standard comparator in all risk prediction research; (3) result in comparable data collection; and (4) allow for individual patient data meta-analyses. This should lead to continued progress in cardiovascular clinical risk prediction. A review of the current evidence suggests that to improve the preoperative clinical risk stratification for adverse cardiac events, a new risk stratification model be built that maintains the clinical risk factors identified in the RCRI, with the following modifications: (1) additional glomerular filtration rate cut points (as opposed to a single creatinine cut point); (2) age; (3) a history of peripheral vascular disease; (4) functional capacity; and (5) a specific surgical procedural category. One would expect a substantial improvement in the discrimination of the RCRI with this approach. Although most noncardiac surgeries will benefit from a standard "generic" cardiovascular risk prediction model, there are data to suggest that patients with human immunodeficiency virus disease who are undergoing vascular surgery may benefit from specific cardiovascular risk prediction models.

How to derive and validate clinical prediction models for use in intensive care medicine

BACKGROUND

Clinical prediction models are formal combinations of historical, physical examination and laboratory or radiographic test data elements designed to accurately estimate the probability that a specific illness is present (diagnostic model), will respond to a form of treatment (therapeutic model) or will have a well-defined outcome (prognostic model) in an individual patient. They are derived and validated using empirical data and used to assist physicians in their clinical decision-making that requires a quantitative assessment of diagnostic, therapeutic or prognostic probabilities at the bedside.

PURPOSE

To provide intensivists with a comprehensive overview of the empirical development and testing phases that a clinical prediction model must satisfy before its implementation into clinical practice.

RESULTS

The development of a clinical prediction model encompasses three consecutive phases, namely derivation, (external) validation and impact analysis. The derivation phase consists of building a multivariable model, estimating its apparent predictive performance in terms of both calibration and discrimination, and assessing the potential for statistical over-fitting using internal validation techniques (i.e. split-sampling, cross-validation or bootstrapping). External validation consists of testing the predictive performance of a model by assessing its calibration and discrimination in different but plausibly related patients. Impact analysis involves comparative research [i.e. (cluster) randomized trials] to determine whether clinical use of a prediction model affects physician practices, patient outcomes or the cost of healthcare delivery.

CONCLUSIONS

This narrative review introduces a checklist of 19 items designed to help intensivists develop and transparently report valid clinical prediction models.