Predicting hospital admissions from individual patient data (IPD): an applied example to explore key elements driving external validity.

OBJECTIVE: To explore factors that potentially impact external validation performance while developing and validating a prognostic model for hospital admissions (HAs) in complex older general practice patients. STUDY DESIGN AND SETTING: Using individual participant data from four cluster-randomised trials conducted in the Netherlands and Germany, we used logistic regression to develop a prognostic model to predict all-cause HAs within a 6-month follow-up period. A stratified intercept was used to account for heterogeneity in baseline risk between the studies. The model was validated both internally and by using internal-external cross-validation (IECV). RESULTS: Prior HAs, physical components of the health-related quality of life comorbidity index, and medication-related variables were used in the final model. While achieving moderate discriminatory performance, internal bootstrap validation revealed a pronounced risk of overfitting. The results of the IECV, in which calibration was highly variable even after accounting for between-study heterogeneity, agreed with this finding. Heterogeneity was equally reflected in differing baseline risk, predictor effects and absolute risk predictions. CONCLUSIONS: Predictor effect heterogeneity and differing baseline risk can explain the limited external performance of HA prediction models. With such drivers known, model adjustments in external validation settings (eg, intercept recalibration, complete updating) can be applied more purposefully. TRIAL REGISTRATION NUMBER: PROSPERO id: CRD42018088129.

Predicting negative health outcomes in older general practice patients with chronic illness: Rationale and development of the PROPERmed harmonized individual participant data database.

The prevalence of multimorbidity and polypharmacy increases significantly with age and are associated with negative health consequences. However, most current interventions to optimize medication have failed to show significant effects on patient-relevant outcomes. This may be due to ineffectiveness of interventions themselves but may also reflect other factors: insufficient sample sizes, heterogeneity of population. To address this issue, the international PROPERmed collaboration was set up to obtain/synthesize individual participant data (IPD) from five cluster-randomized trials. The trials took place in Germany and The Netherlands and aimed to optimize medication in older general practice patients with chronic illness. PROPERmed is the first database of IPD to be drawn from multiple trials in this patient population and setting. It offers the opportunity to derive prognostic models with increased statistical power for prediction of patient-relevant outcomes resulting from the interplay of multimorbidity and polypharmacy. This may help patients from this heterogeneous group to be stratified according to risk and enable clinicians to identify patients that are likely to benefit most from resource/time-intensive interventions. The aim of this manuscript is to describe the rationale behind PROPERmed collaboration, characteristics of the included studies/participants, development of the harmonized IPD database and challenges faced during this process.

A prognostic model predicted deterioration in health-related quality of life in older patients with multimorbidity and polypharmacy.

OBJECTIVES: To develop and validate a prognostic model to predict deterioration in health-related quality of life (dHRQoL) in older general practice patients with at least one chronic condition and one chronic prescription. STUDY DESIGN AND SETTING: We used individual participant data from five cluster-randomized trials conducted in the Netherlands and Germany to predict dHRQoL, defined as a decrease in EQ-5D-3 L index score of ≥5% after 6-month follow-up in logistic regression models with stratified intercepts to account for between-study heterogeneity. The model was validated internally and by using internal-external cross-validation (IECV). RESULTS: In 3,582 patients with complete data, of whom 1,046 (29.2%) showed deterioration in HRQoL, and 12/87 variables were selected that were related to single (chronic) conditions, inappropriate medication, medication underuse, functional status, well-being, and HRQoL. Bootstrap internal validation showed a C-statistic of 0.71 (0.69 to 0.72) and a calibration slope of 0.88 (0.78 to 0.98). In the IECV loop, the model provided a pooled C-statistic of 0.68 (0.65 to 0.70) and calibration-in-the-large of 0 (-0.13 to 0.13). HRQoL/functionality had the strongest prognostic value. CONCLUSION: The model performed well in terms of discrimination, calibration, and generalizability and might help clinicians identify older patients at high risk of dHRQoL. REGISTRATION: PROSPERO ID: CRD42018088129.

Urinary excretion of biomarkers of oxidatively damaged DNA and RNA in hereditary hemochromatosis.

Oxidatively generated damage to nucleic acids is considered to play a significant role in carcinogenesis, and it has been shown that people with hereditary hemochromatosis are at increased risk of cancer. In this study we used a new refined liquid chromatography-tandem mass spectrometry method to measure the urinary excretion of oxidatively generated 8-oxo-7,8-dihydroguanine and related 2'-deoxyribonucleoside and ribonucleoside derivatives in hereditary hemochromatosis patients, and we investigated the effect of treatment on the levels of these modifications. The study was carried out as a classical case-control study of 21 newly diagnosed, never treated hereditary hemochromatosis patients and 21 matched controls. We found that at baseline the urinary excretion of the RNA oxidation product 8-oxo-7,8-dihydroguanosine (8-oxoGuo) was 2.5-fold increased in patients compared with controls, and after phlebotomy treatment the excretion of the RNA oxidation product 8-oxoGuo returned to control values and the excretion of the DNA product 8-oxo-7,8-dihydro-2'-deoxyguanosine was reduced by 30%. In patients with hereditary hemochromatosis oxidative stress on nucleic acids is an important feature of the iron overload seen in this disease. By this mechanism cellular damage resulting in end organ damage, typically seen in the liver of such patients, may be mediated.

Correlation of biomarkers of endothelium dysfunction and matrix remodeling in patients with systemic sclerosis.

OBJECTIVE: Systemic sclerosis (SSc) is a multisystem disease characterized by microvascular dysfunction and excessive fibrosis. However, the relationship between these 2 features remains unclear. Endothelial dysfunction can be assessed by quantifying plasma asymmetric dimethylarginine (ADMA), an endogenous inhibitor of endothelial nitric oxide synthase. Matrix remodeling can be assessed by quantifying serum tissue inhibitor of matrix metalloproteinases-1 (TIMP-1). Both biomarkers are elevated in patients with SSc. Our objective was to test whether plasma ADMA is correlated with serum TIMP-1. METHODS: We enrolled 91 subjects, 39 patients with SSc, 28 patients with primary Raynaud's phenomenon (RP), and 24 healthy volunteers. Plasma ADMA concentrations were measured by liquid chromatography-tandem mass spectrometry. Serum TIMP-1 concentrations were determined by ELISA. RESULTS: Mean ADMA concentrations were higher in patients with SSc (0.68 microM +/- 0.12) than in patients with primary RP or healthy volunteers (respectively, 0.56 microM +/- 0.14 and 0.62 microM +/- 0.12; p = 0.002). Median serum TIMP-1 concentrations were increased in patients with SSc compared to primary RP and healthy volunteers [12 (9-15), 11 (8-13), and 10 (7-13) nM, respectively; p = 0.05]. In the SSc group, we observed a statistically significant correlation between plasma ADMA and serum TIMP-1 (r = 0.34, p = 0.035). CONCLUSION: These data are consistent with our hypothesis of an association of endothelial dysfunction and matrix remodeling in scleroderma spectrum disorders.

Isoprostanes inhibit vascular endothelial growth factor-induced endothelial cell migration, tube formation, and cardiac vessel sprouting in vitro, as well as angiogenesis in vivo via activation of the thromboxane A(2) receptor: a potential link between oxidative stress and impaired angiogenesis.

Isoprostanes are endogenously formed end products of lipid peroxidation. Furthermore, they are markers of oxidative stress and independent risk markers of coronary heart disease. In patients experiencing coronary heart disease, impaired angiogenesis may exacerbate insufficient blood supply of ischemic myocardium. We therefore hypothesized that isoprostanes may exert detrimental cardiovascular effects by inhibiting angiogenesis. We studied the effect of isoprostanes on vascular endothelial growth factor (VEGF)-induced migration and tube formation of human endothelial cells (ECs), and cardiac angiogenesis in vitro as well as on VEGF-induced angiogenesis in the chorioallantoic membrane assay in vivo. The isoprostanes 8-iso-PGF(2alpha), 8-iso-PGE(2), and 8-iso-PGA(2) inhibited VEGF-induced migration, tube formation of ECs, and cardiac angiogenesis in vitro, as well as VEGF-induced angiogenesis in vivo via activation of the thromboxane A(2) receptor (TBXA2R): the specific TBXA2R antagonists SQ-29548, BM 567, and ICI 192,605 but not the thromboxane A(2) synthase inhibitor ozagrel blocked the effect of isoprostanes. The isoprostane 8-iso-PGA(2) degraded into 2 biologically active derivatives in vitro, which also inhibited EC tube formation via the TBXA2R. Moreover, short hairpin RNA-mediated knockdown of the TBXA2R antagonized isoprostane-induced effects. In addition, Rho kinase inhibitor Y-27632 reversed the inhibitory effect of isoprostanes and the thromboxane A(2) mimetic U-46619 on EC migration and tube formation. Finally, the various isoprostanes exerted a synergistic inhibitory effect on EC tube formation. We demonstrate for the first time that isoprostanes inhibit angiogenesis via activation of the TBXA2R. By this mechanism, isoprostanes may contribute directly to exacerbation of coronary heart disease and to capillary rarefaction in disease states of increased oxidative stress.

Impact of atorvastatin treatment on platelet-activating factor acetylhydrolase and 15-F(2trans)-isoprostane in hypercholesterolaemic patients.

AIMS: Isoprostanes are the product of free radical oxidation of arachidonic acid, whose hydrolysis from phospholipids is presumably catalysed by phospholipases A(2) (PLA(2)s) such as group IIA or V PLA(2)s, or group VII PLA(2)[platelet-activating factor acetylhydrolase (PAF-AH), lipoprotein-associated phospholipase]. Atorvastatin reduces concentrations of low-density lipoprotein (LDL), with which PAF-AH is associated, and PLA(2)s' protein concentrations. We investigated the effect of atorvastatin on PLA(2)s and PAF-AH activity and the urinary excretion of 15-F(2trans)-isoprostane (15-F(2t)-IsoP, 8-iso-PGF(2alpha), iPF(2alpha)-III). METHODS: Twenty-four hypercholesterolaemic individuals naive to lipid-lowering therapy were randomized to atorvastatin 40 mg or placebo for 6 weeks. The 15-F(2t)-isoP urinary excretion (gas chromatography/mass spectrometry), PAF-AH and group IIA and V PLA(2) activities (photometry) were assessed at baseline and end-point. RESULTS: At end-point, 15-F(2t)-isoP urinary excretion concentrations as well as PLA(2)s' activity were unchanged under atorvastatin (mean change 0.21 +/- 1.79 ng h(-1), 95% confidence interval -0.92, 1.35 and 0.33 +/- 0.94 nmol min(-1) ml(-1), -0.27, 0.93) and under placebo (mean change 0.69 +/- 1.69 ng h(-1), -0.52, 1.90 and 1.29 +/- 2.16 nmol min(-1) ml(-1), -0.25, 2.84). Atorvastatin treatment decreased total (P < 0.001) and LDL-cholesterol (P < 0.001) but had no effect on high-density lipoprotein. PAF-AH activity was lowered in the atorvastatin group (mean change - 5.27+/- 1.96 nmol min(-1) ml(-1), -6.51, -4.03, P < 0.001) but not in the placebo group (mean change 1.02 +/- 1.64 nmol min(-1) ml(-1), 0.15, 2.20), and the change in PAF-AH activity was correlated with that in total (P = 0.03) and LDL-cholesterol (P = 0.03). CONCLUSION: Our results show a lowering effect of atorvastatin on PAF-AH activity associated with its lipid-lowering effect and exclude a key role of PAF-AH in the liberation of 15-F(2t)-isoP from phospholipids.

Postocclusive reactive hyperemia inversely correlates with urinary 15-F2t-isoprostane levels in systemic sclerosis.

Microvascular dysfunction and increased oxidative stress are major hallmarks of the systemic sclerosis disease process. The primary objective of this study was to test whether there is a link between peak postocclusive hyperemia and urinary levels of the F2-isoprostane 15-F2t-IsoP (8-iso-PGF2alpha) in patients suffering from systemic sclerosis. We enrolled 43 patients suffering from systemic sclerosis, 33 patients with primary Raynaud's phenomenon (RP), and 25 healthy volunteers. Microvascular function was assessed using the postocclusive hyperemia monitored by laser Doppler flowmetry. Endothelium-independent response was monitored after 0.4 mg sublingual nitroglycerin. Oxidative stress status was assessed by urinary levels of the F2-isoprostane 15-F2t-IsoP using GC-MS. The peak postocclusive vascular conductance was altered in subjects with systemic sclerosis and primary RP compared to controls (respectively 28 (7-48), 30 (13-48), and 39.9 (13-63) mV/mm Hg, p = 0.01). F2-isoprostanes were increased in the systemic sclerosis group compared to primary Raynaud's phenomenon and healthy controls (respectively 230 (155-387), 182 (101-284), and 207 (109-291) pg/mg, p = 0.006). In patients suffering from systemic sclerosis, there was a significant inverse correlation between F2-isoprostanes and postocclusive hyperemia, expressed as raw data (R = -0.45, p = 0.007) or as an increase over baseline (R = -0.28, p = 0.04). Conversely, no correlation was found with the nitroglycerin response. In conclusion, we provide evidence that there is an inverse correlation between postocclusive hyperemia and urinary F2-isoprostane levels in patients suffering from systemic sclerosis. Whether oxygen free radicals initiate the vascular dysfunction or whether there is an initial trigger that initiates both processes will need to be further clarified in future studies.