Flexible Parametric Survival Models Research Articles

Concordance of Immunohistochemistry-Based and Gene Expression-Based Subtyping in Breast Cancer.

BackgroundUse of immunohistochemistry-based surrogates of molecular breast cancer subtypes is common in research and clinical practice, but information on their comparative validity and prognostic capacity is scarce.MethodsData from 2 PAM50-subtyped Swedish breast cancer cohorts were used: Stockholm tamoxifen trial–3 with 561 patients diagnosed 1976-1990 and Clinseq with 237 patients diagnosed 2005-2012. We evaluated 3 surrogate classifications; the immunohistochemistry-3 surrogate classifier based on estrogen receptor, progesterone receptor, and HER2 and the St. Gallen and Prolif surrogate classifiers also including Ki-67. Accuracy, kappa, sensitivity, and specificity were computed as compared with PAM50. Alluvial diagrams of misclassification patterns were plotted. Distant recurrence-free survival was assessed using Kaplan-Meier plots, and tamoxifen treatment benefit for luminal subtypes was modeled using flexible parametric survival models.ResultsThe concordance with PAM50 ranged from poor to moderate (kappa = 0.36-0.57, accuracy = 0.54-0.75), with best performance for the Prolif surrogate classification in both cohorts. Good concordance was only achieved when luminal subgroups were collapsed (kappa = 0.71-0.69, accuracy = 0.90-0.91). The St. Gallen surrogate classification misclassified luminal A into luminal B; the reverse pattern was seen with the others. In distant recurrence-free survival, surrogates were more similar to each other than PAM50. The difference in tamoxifen treatment benefit between luminal A and B for PAM50 was not replicated with any surrogate classifier.ConclusionsAll surrogate classifiers had limited ability to distinguish between PAM50 luminal A and B, but patterns of misclassifications differed. PAM50 subtyping appeared to yield larger separation of survival between luminal subtypes than any of the surrogate classifications.

Effect of Lipid Abnormality on CKD Progression from Moderate to Severe Stage: Application of Flexible Parametric Proportional-Hazards and Proportional-Odds Models

Background: Lipid disorders are a well-documented risk factor for chronic kidney disease (CKD), but the impact of lipid abnormalities in the progression of the disease remains mixed. Objectives: The current study aimed to extend the existing knowledge about the effect of lipid disorders in disease progression from moderate to severe stage using Flexible parametric survival models. Methods: This retrospective cohort study included 308 moderate CKD patients who received the nephrologist follow-up visits at the nephrology clinic, Ilam (Iran), from 2012 to 2019. The survival time was determined based on the time medically diagnosed with moderate stages (GFR = 59 - 55 mL/min per 1.73 m2) to the time of progression to the severe stage (GFR = 29 - 25 mL/min per 1.73 m2) hazard using flexible parametric survival models. Results: In univariate analysis, high levels of TG, LDL, and cholesterol were important risk factors which affect the CKD progression. The hazard of patients with TG > 200 mg/dL was 1.69 times higher than patients with desirable TG levels (P = 0.09). Moreover, for patients with LDL > 160 mg/dL, the hazard was 2.12 times higher than patients with desirable LDL levels (P = 0.01). The hazard of patients with total cholesterol levels > 240 mg/dL was 2.10 times higher than patients with desirable cholesterol levels (P = 0.003). The adjusted model was shown to better fit the PH model. Cholesterol levels > 240 mg/dL remains a significant risk factor for CKD progression (P = 0.03). Conclusions: Effective treatment programs should pay closer attention to screening and treatment of hyperlipidemia in patients diagnosed with moderate CKD.

Trends in short-term survival from distant-stage cutaneous melanoma in the United States, 2001-2013 (CONCORD-3).

BackgroundSurvival from metastatic cutaneous melanoma is substantially lower than for localized disease. Treatments for metastatic melanoma have been limited, but remarkable clinical improvements have been reported in clinical trials in the last decade. We described the characteristics of US patients diagnosed with cutaneous melanoma during 2001-2013 and assessed trends in short-term survival for distant-stage disease.MethodsTrends in 1-year net survival were estimated using the Pohar Perme estimator, controlling for background mortality with life tables of all-cause mortality rates by county of residence, single year of age, sex, and race for each year 2001-2013. We fitted a flexible parametric survival model on the log-hazard scale to estimate the effect of race on the hazard of death because of melanoma and estimated 1-year net survival by race.ResultsOnly 4.4% of the 425 915 melanomas were diagnosed at a distant stage, cases diagnosed at a distant stage are more commonly men, older patients, and African Americans. Age-standardized, 1-year net survival for distant-stage disease was stable at approximately 43% during 2001-2010. From 2010 onward, survival improved rapidly, reaching 58.9% (95% confidence interval = 56.6% to 61.2%) for patients diagnosed in 2013. Younger patients experienced the largest improvement. Survival for distant-stage disease increased in both Blacks and Whites but was consistently lower in Blacks.ConclusionsOne-year survival for distant-stage melanoma improved during 2001-2013, particularly in younger patients and those diagnosed since 2010. This improvement may be a consequence of the introduction of immune-checkpoint-inhibitors and other targeted treatments for metastatic and unresectable disease. Persistent survival inequalities exist between Blacks and Whites, suggesting differential access to treatment.

Impact of area-level socioeconomic status and accessibility to treatment on life expectancy after a cancer diagnosis in Queensland, Australia

AimsThis study quantifies geographic inequities in loss of life expectancy (LOLE) by area-level socioeconomic status (SES) and accessibility to treatment. MethodsAnalysis was conducted using a population-based cancer-registry cohort (n = 371,570) of Queensland (Australia) residents aged 50−89 years, diagnosed between 1997–2016. Flexible parametric survival models were used to estimate LOLE by area-level SES and accessibility for all invasive cancers and the five leading cancers. The gain in life years that could be achieved if all cancer patients experienced the same relative survival as those in the least disadvantaged-high accessibility category was estimated for the 2016 cohort. ResultsFor all invasive cancers, men living in the most disadvantaged areas lost 34 % of life expectancy due to their cancer diagnosis, while those from the least disadvantaged areas lost 25 %. The corresponding percentages for women were 33 % and 23 %. Accessibility had a lower impact on LOLE than SES, with patients from low accessibility areas losing 0–4 % more life expectancy than those from high accessibility areas. For cancer patients diagnosed in 2016 (n = 24,423), an estimated 101,387 life years will be lost. This would be reduced by 19 % if all patients experienced the same relative survival as those from the least disadvantaged-high accessibility areas. ConclusionThe impact of a cancer diagnosis on remaining life expectancy varies by geographical area. Establishing reasons why area disadvantage impacts on life expectancy is crucial to inform subsequent interventions that could increase the life expectancy of cancer patients from more disadvantaged areas.

Patients Receiving a Primary Unicompartmental Knee Replacement Have a Higher Risk of Revision but a Lower Risk of Mortality Than Predicted Had They Received a Total Knee Replacement: Data From the National Joint Registry for England, Wales, Northern Ireland, and the Isle of Man

BackgroundTo determine unicompartmental (UKR) and total knee replacement (TKR) revision rates, compare UKR revision rates with what they would have been had they received TKR instead, and assess subsequent re-revision and 90-day mortality rates. MethodsUsing National Joint Registry data, we estimated UKR and TKR revision and mortality rates. Flexible parametric survival modeling (FPM) was used to model failure in TKR and make estimates for UKR. Kaplan-Meier estimates were used to compare cumulative re-revision for revised UKRs and TKRs. ResultsTen-year UKR revision rates were 2.5 times higher than expected from TKR, equivalent to 70 excess revisions/1000 cases within 10 years (5861 excess revisions in this cohort). Revision rates were 2.5 times higher for the highest quartile volume UKR surgeons compared to the same quartile for TKR and 3.9 times higher for the lowest quartiles respectively. Re-revision rates of revised TKRs (10 years = 17.5%, 95% confidence interval [CI] 16.4-18.7) were similar to revised UKRs (15.2%, 95% CI 13.4-17.1) and higher than revision rates following primary TKR (3.3%, 95% CI 3.1-3.5). Ninety-day mortality rates were lower after UKR compared with TKR (0.08% vs 0.33%) and lower than predicted had UKR patients received a TKR (0.18%), equivalent to 1 fewer death per 1000 cases. ConclusionUKR revision rates were substantially higher than TKR even when demographics and caseload differences were accounted for; however, fewer deaths occur after UKR. This should be considered when forming treatment guidelines and commissioning services. Re-revision rates were similar between revised UKRs and TKRs, but considerably higher than for primary TKR, therefore UKR cannot be considered an intermediate procedure.

Age-dependency in mortality of Finnish family caregivers: a nationwide register-based study

Abstract Purpose Mortality appears to be lower in family caregivers than in the general population. However, there is lack of knowledge whether the difference in mortality between family caregivers and the general population is dependent on age. The purpose of this study was to analyze all-cause mortality in relation to age in family caregivers and to study their cause-specific mortality using data from multiple Finnish national registers. Methods The data included all individuals, who received family caregiver's allowance in Finland in 2012 (n = 42 256, mean age 67 years, 71% women) and a control population matched for age, sex, and municipality of residence (n = 83 618). Information on dates and causes of death between 2012 and 2017 were obtained from the Finnish Causes of Death Register. Flexible parametric survival modeling and competing risk regression adjusted for socioeconomic status were used. Results The total follow-up time was 717 877 person-years. Family caregivers had lower all-cause mortality than the controls over the follow-up (8.1% vs. 11.6%) both among women (hazard ratio [HR]: 0.64, 95% CI: 0.61-0.68) and men (HR: 0.73, 95% CI: 0.70-0.77). Younger adult caregivers had equal or only slightly lower mortality than their controls, but after age 60, the difference increased markedly resulting in over 10% lower mortality in favor of the caregivers in the oldest age groups. Caregivers had lower mortality for all the causes of death studied, namely cardiovascular, cancer, neurological, external, respiratory, gastrointestinal and dementia than the controls. Of these, the lowest was the risk for dementia (subhazard ratio=0.29, 95%CI: 0.25-0.34). Conclusions Older family caregivers have lower mortality than the age-matched controls from the general population while younger caregivers have similar mortality to their peers. This age-dependent advantage in mortality is likely to reflect selection of healthier individuals into the family caregiver role. Key messages The difference in mortality between family caregivers and the age-matched general population varies considerably with age. Advantage in mortality observed in family caregiver studies is likely to reflect the selection of healthier individuals into the caregiver role, which underestimates the adverse effects of caregiving.

Quantifying the Number of Cancer Deaths Avoided Due to Improvements in Cancer Survival since the 1980s in the Australian Population, 1985-2014.

This study quantifies the number of potentially "avoided"cancer deaths due to differences in 10-year relative survival between three time periods, reflecting temporal improvements in cancer diagnostic and/or treatment practices in Australia. National population-based cohort of 2,307,565 Australians ages 15 to 89 years, diagnosed with a primary invasive cancer from 1985 to 2014 with mortality follow-up to December 31, 2015. Excess mortality rates and crude probabilities of cancer deaths were estimated using flexible parametric relative survival models. Crude probabilities were then used to calculate "avoided cancer deaths" (reduced number of cancer deaths within 10 years of diagnosis due to survival changes since 1985-1994) for all cancers and 13 leading cancer types. For each cancer type, excess mortality (in the cancer cohort vs. the expected population mortality) was significantly lower for more recently diagnosed persons. For all cancers combined, the number of "avoided cancer deaths" (vs. 1985-1994) was 4,877 (1995-2004) and 11,385 (2005-2014) among males. Prostate (1995-2004: 2,144; 2005-2014: 5,099) and female breast cancer (1,127 and 2,048) had the highest number of such deaths, whereas <400 were avoided for pancreatic or lung cancers across each period. Screening and early detection likely contributed to the high number of "avoided cancer deaths" for prostate and female breast cancer, whereas early detection remains difficult for lung and pancreatic cancers, highlighting the need for improved preventive and screening measures. Absolute measures such as "avoided cancer deaths" can provide a more tangible estimate of the improvements in cancer survival than standard net survival measures.

The effect of surgical approach in total hip replacement on outcomes: an analysis of 723,904 elective operations from the National Joint Registry for England, Wales, Northern Ireland and the Isle of Man

BackgroundTotal hip replacement (THR) is clinically and cost-effective. The surgical approach employed influences the outcome; however, there is little generalisable and robust evidence to guide practice.MethodsA total of 723,904 primary THRs captured in the National Joint Registry, linked to hospital inpatient, mortality and patient-reported outcome measures (PROMs) data with up to 13.75 years follow-up, were analysed. There were seven surgical approach groups: conventional posterior, lateral, anterior and trans-trochanteric groups and minimally invasive posterior, lateral and anterior. Survival methods were used to compare revision rates and 90-day mortality. Groups were compared using Cox proportional hazards and Flexible Parametric Survival Modelling (FPM). Confounders included age at surgery, sex, risk group (indications additional to osteoarthritis), American Society of Anesthesiologists grade, THR fixation, thromboprophylaxis, anaesthetic, body mass index (BMI) and deprivation. PROMs were analysed with regression modelling or non-parametric methods.ResultsUnadjusted analysis showed a higher revision risk than the referent conventional posterior for the conventional lateral, minimally invasive lateral, minimally invasive anterior and trans-trochanteric groups. This persisted with all adjusted FPM and adjusted Cox models, except in the Cox model including BMI where the higher revision rate only persisted for the conventional lateral approach (hazard rate ratio (HRR) 1.12 [95% CI 1.06,1.17] P < 0·001) and trans-trochanteric approaches (HRR 1.48 [95% CI 1.14,1.91] P = 0.003). PROMs demonstrated statistically, but not clinically, significant differences. Self-reported complications were more frequent with the conventional lateral approach, and the risk of 90-day mortality was higher (HRR 1.15 [95%CI 1.01–1.30] P = 0.029).ConclusionsLateral approaches for THR are associated with worse outcomes, including more deaths and revisions, than the posterior approach. We recommend the posterior approach should be considered the current standard approach for THR. Large well-designed studies are needed to assess any potential benefits from using minimally invasive posterior approaches and the conventional anterior approach.

Antipsychotic Drugs and Hip Fracture: Associations Before and After the Initiation of Treatment

ObjectiveTo study the association between antipsychotic drug treatment and hip fracture, before and after the initiation of treatment. DesignNationwide cohort study. Setting and ParticipantsIn this study based on several Swedish registers, all individuals age ≥65 years who filled prescriptions for antipsychotic drugs in 2007–2017 were matched 1:1 by sex and age with controls, resulting in a cohort of 255,274 individuals. MeasuresAssociations between antipsychotic drug treatment and hip fracture were investigated using multivariable conditional logistic regression models and flexible parametric survival models for nonproportional hazards, starting 1 year before the first prescription was filled and extending to 1 year thereafter. ResultsThe studied cohort had a mean age of 81.5 (standard deviation, 8.1) years; 152,890 (59.9%) individuals were women. Antipsychotic drug use was associated with an increased risk of hip fracture in all studied time frames, before and after the initiation of treatment. The risk was highest 16–30 days before the initiation of treatment (odds ratio 9.09; 95% confidence interval 7.00–11.81). The pattern was consistent in subgroup analyses of users of conventional and atypical antipsychotics, men and women, as well as in younger old and older old participants. The association with hip fracture was not influenced by antipsychotic drug dose. Conclusions and ImplicationsThe association between antipsychotic drug use and the risk of hip fracture was observed before the initiation of antipsychotic treatment. This finding suggests that factors other than exposure to antipsychotic drugs are responsible for the increased risk of hip fracture in the treatment group.

Associations between Lesion Locations and Stroke Recurrence in Survivors of First-ever Ischemic Stroke: A Prospective Cohort Study.

Several studies have indicated that stroke survivors with multiple lesions or with larger lesion volumes have a higher risk of stroke recurrence. However, the relationship between lesion locations and stroke recurrence is unclear. We conducted a prospective cohort study of first-ever ischemic stroke survivors who were consecutively enrolled from January 2010 to December 2015. Stroke recurrence was assessed every 3 months after post-discharge via telephone interviews by trained interviewers. Lesion locations were obtained from hospital-based MRI or CT scans and classified using two classification systems that were based on cerebral hemisphere or vascular territory and brain anatomical structures. Flexible parametric survival models using the proportional hazards scale (PH model) were used to analyze the time-to-event data. Among 633 survivors, 63.51% (n=402) had anterior circulation ischemia (ACI), and more than half of all ACIs occurred in the subcortex. After a median follow-up of 2.5 years, 117 (18.48%) survivors developed a recurrent stroke. The results of the multivariate PH model showed that survivors with non-brain lesions were at higher risk of recurrence than those with right-side lesions (HR, 2.79; 95%CI, 1.53, 5.08; P=0.001). There was no increase in risk among survivors with left-side lesions (HR, 0.97; 95%CI, 0.53, 1.75; P=0.914) or both-side lesions (HR, 1.24; 95%CI, 0.75, 2.07; P=0.401) compared to those with right-side lesions. Additionally, there were no associations between stroke recurrence and lesion locations that were classified based on vascular territory and brain anatomical structures. It was concluded that first-ever ischemic stroke survivors with non-brain lesion had higher recurrence risk than those with right-side lesion, although no significant associations were found when the lesion locations were classified by vascular territory and brain anatomical structures.

Associations of blood pressure with risk of injurious falls in old age vary by functional status: A cohort study.

We aimed to examine to what extent blood pressure (BP) components are associated with injurious falls in older adults, and whether the possible associations differ by functional status (indicated by physical and cognitive impairment at baseline). This prospective cohort study included 3055 community-living participants of the Swedish National study on Aging and Care in Kungsholmen (aged ≥60years). At baseline (2001-2004), we measured systolic BP (SBP) and diastolic BP (DBP), mean arterial pressure, pulse pressure (PP), and orthostatic hypotension. Physical function was assessed using tests of balance, chair stands, and walking speed. Cognitive function was assessed with the Mini-Mental State Examination. Injurious falls leading to inpatient or outpatient care during 3 and 10years of follow-up were identified via patient registers. Data were analyzed using flexible parametric survival models adjusted for potential confounders. During the 10-year follow-up period in people without functional impairment, the multi-adjusted hazard ratios (HR) of injurious falls were 1.77 (95% CI 1.02-3.07) for having SBP <130mmHg, 1.73 (95% CI 1.05-2.83) for having SBP ≥160mmHg (vs. 130-139), and 1.46 (1.05-2.02) for having higher tertile of PP (vs. lower tertile). During the 3-year follow-up period in people with functional impairment, the multi-adjusted HR of injurious falls was 1.91 (95% CI 1.17-3.13) for having SBP <130mmHg (vs. 130-139) and 0.74 (0.59-0.94) for having higher tertile of PP (vs. lower tertile). There was no significant association between BP components and injurious falls either in people without functional impairment during the 3-year follow-up period, or in people with functional impairment during the 10-year follow-up period. This large-scale Swedish study provides evidence that the associations between some specific components of BP and risk of injurious falls in older adults vary by functional status. This supports the view that a personalized approach to blood pressure management depending on functional status among older adults might be warranted in clinical settings.

Coffee consumption and overall and cause-specific mortality: the Norwegian Women and Cancer Study (NOWAC)

Coffee consumption has previously been reported to reduce overall and cause-specific mortality. We aimed to further investigate this association by coffee brewing methods and in a population with heavy coffee consumers. The information on total, filtered, instant, and boiled coffee consumption from self-administered questionnaires was available from 117,228 women in the Norwegian Women and Cancer (NOWAC) Study. We used flexible parametric survival models to calculate hazard ratios (HR) and 95% confidence intervals (CI) for all-cause, cardiovascular, and cancer mortality by total coffee consumption and brewing methods, and adjusted for smoking status, number of pack-years, age at smoking initiation, alcohol consumption, body mass index, physical activity, and duration of education. During 3.2 million person-years of follow-up, a total of 16,106 deaths occurred. Compared to light coffee consumers (≤ 1 cup/day), we found a statistically significant inverse association with high-moderate total coffee consumption (more than 4 and up to 6 cups/day, HR 0.89; 95% CI 0.83–0.94) and all-cause mortality. The adverse association between heavy filtered coffee consumption (> 6 cups/day) and all-cause mortality observed in the entire sample (HR 1.09; 95% CI 1.01–1.17) was not found in never smokers (HR 0.85; 95% CI 0.70–1.05). During the follow-up, both high-moderate total and filtered coffee consumption were inversely associated with the risk of cardiovascular mortality (HR 0.79; 95% CI 0.67–0.94; HR 0.80; 95% CI 0.67–0.94, respectively). The association was stronger in the analyses of never smokers (> 6 cups of filtered coffee/day HR 0.20; 95% CI 0.08–0.56). The consumption of more than 6 cups/day of filtered, instant, and coffee overall was found to increase the risk of cancer deaths during the follow-up. However, these associations were not statistically significant in the subgroup analyses of never smokers. The data from the NOWAC study indicate that the consumption of filtered coffee reduces the risk of cardiovascular deaths. The observed adverse association between coffee consumption and cancer mortality is most likely due to residual confounding by smoking.

Attrition from specialised rehabilitation associated with an elevated mortality risk: results from a vital status tracing study in Swiss spinal cord injured patients.

IntroductionStudy drop-out and attrition from treating clinics is common among persons with chronic health conditions. However, if attrition is associated with adverse health outcomes, it may bias or mislead inferences for health policy and resource allocation.MethodsThis retrospective cohort study uses data attained through the Swiss Spinal Cord Injury (SwiSCI) cohort study on persons with spinal cord injury (SCI). Vital status (VS) was ascertained either through clinic medical records (MRs) or through municipalities in a secondary tracing effort. Flexible parametric survival models were used to investigate risk factors for going lost to clinic (LTC) and the association of LTC with subsequent risk of mortality.Results1924 individuals were included in the tracing study; for 1608 of these cases, contemporary VS was initially checked in the MRs. VS was ascertained for 704 cases of the 1608 cases initially checked in MRs; of the remaining cases (n=904), nearly 90% were identified in municipalities (n=804). LTC was associated with a nearly fourfold higher risk of mortality (HR=3.62; 95% CI 2.18 to 6.02) among persons with traumatic SCI. Extended driving time (ie, less than 30 min compared with 30 min and longer to reach the nearest specialised rehabilitation facility) was associated with an increased risk of mortality (HR=1.51, 95% CI 1.02 to 2.22) for individuals with non-traumatic SCI.ConclusionThe differential risk of LTC according to sociodemographic and SCI lesion characteristics underscores the importance of accounting for attrition in cohort studies on chronic disease populations requiring long-term care. In addition, given the associated risk of mortality, LTC is an issue of concern to clinicians and policy makers aiming to optimise the long-term survival of community-dwelling individuals with traumatic SCI. Future studies are necessary to verify whether it is possible to improve survival prospects of individuals LTC through more persistent outreach and targeted care.

Right single lung transplantation or double lung transplantation compared with left single lung transplantation in chronic obstructive pulmonary disease

Although single and double lung transplantation outcomes for chronic obstructive pulmonary disease (COPD) have been investigated, right and left single lung transplants have never been rigorously compared to evaluate disease-specific differences. Single lung transplants for COPD often have hyperinflation of the contralateral native lung, which may be more pronounced in left lung transplants. Using the United Network for Organ Sharing registry, we conducted a retrospective cohort study of 5,585 adults who underwent lung transplantation for COPD from May 4, 2005 to June 30, 2017. Subjects were followed until March 2019. Post-transplant survival was compared using Cox proportional hazards and Royston and Parmar's flexible parametric survival models. We adjusted for donor and recipient factors with known or plausible associations with survival. Lung transplant recipients who received a left single lung transplant for COPD had an increased risk of post-transplant death when compared with those who received a right single lung transplant for COPD (hazard ratio [HR]: 1.24, 95% CI: 1.08-1.48, p = 0.002). Survival did not differ significantly between double lung transplant and right single lung transplant recipients (HR: 0.88, 95% CI: 0.77-1.02, p = 0.086). Adjusted 5-year survival was 57.8% (95% CI: 55.7-60.1) for double lung recipients, 56.7% (95% CI: 55.4-58.0) for right single lung recipients, and 50.9% (95% CI: 47.2-55.0) for left single lung recipients. In COPD, right single lung transplantation was associated with improved post-transplant survival compared with left single lung transplantation, and no significant difference in post-transplant survival compared with double lung transplantation was found. In light of the ongoing donor lung shortage, preferential allocation of right single lungs to patients with COPD should be considered.

The Impact of Cardiovascular Disease and Chronic Kidney Disease on Life Expectancy and Direct Medical Cost in a 10-Year Diabetes Cohort Study.

The relative effects of various cardiovascular diseases (CVDs) and varying severity of chronic kidney disease (CKD) on mortality risk, direct medical cost, and life expectancy in patients with diabetes are unclear. The aim of this study was to evaluate these associations. This was a retrospective cohort study that included 208,792 adults with diabetes stratified into 12 disease status groups with varying combinations of heart disease, stroke, moderate CKD (estimated glomerular filtration rate [eGFR] 30-59 mL/min/1.73 m2) and severe CKD (eGFR <30 mL/min/1.73 m2) in 2008-2010. The effect of risk of mortality, annual direct medical costs, and life expectancy were assessed using Cox regression, gamma generalized linear method with log-link function, and flexible parametric survival models. Over a median follow-up of 8.5 years (1.6 million patient-years), 50,154 deaths were recorded. Mortality risks for patients with only a single condition among heart disease, stroke, and moderate CKD were similar. The mortality risks were 1.75 times, 2.63 times, and 3.58 times greater for patients with one, two, and all three conditions (consisting of stroke, heart disease, and moderate CKD), compared with patients without these diseases, suggesting an independent and individually additive effect for any combination. A similar trend was observed in annual public health care costs with 2.91-, 3.90-, and 3.88-fold increased costs for patients with one, two, and three conditions, respectively. Increases in the number of conditions reduced life expectancy greatly, particularly in younger patients. Reduction in life expectancy for a 40-year-old with one, two, and three conditions was 20, 25, and 30 years for men and 25, 30, and 35 years, respectively, for women. A similar trend of greater magnitude was observed for severe CKD. The effects of heart diseases, stroke, CKD, and the combination of these conditions on all-cause mortality and direct medical costs are independent and cumulative. CKD, especially severe CKD, appears to have a particularly significant impact on life expectancy and direct medical costs in patients with diabetes. These findings support the importance of preventing both CVD and CKD in patients with diabetes.

Temporal recalibration for improving prognostic model development and risk predictions in settings where survival is improving over time.

BackgroundPrognostic models are typically developed in studies covering long time periods. However, if more recent years have seen improvements in survival, then using the full dataset may lead to out-of-date survival predictions. Period analysis addresses this by developing the model in a subset of the data from a recent time window, but results in a reduction of sample size.MethodsWe propose a new approach, called temporal recalibration, to combine the advantages of period analysis and full cohort analysis. This approach develops a model in the entire dataset and then recalibrates the baseline survival using a period analysis sample. The approaches are demonstrated utilizing a prognostic model in colon cancer built using both Cox proportional hazards and flexible parametric survival models with data from 1996–2005 from the Surveillance, Epidemiology, and End Results (SEER) Program database. Comparison of model predictions with observed survival estimates were made for new patients subsequently diagnosed in 2006 and followed-up until 2015.ResultsPeriod analysis and temporal recalibration provided more up-to-date survival predictions that more closely matched observed survival in subsequent data than the standard full cohort models. In addition, temporal recalibration provided more precise estimates of predictor effects.ConclusionPrognostic models are typically developed using a full cohort analysis that can result in out-of-date long-term survival estimates when survival has improved in recent years. Temporal recalibration is a simple method to address this, which can be used when developing and updating prognostic models to ensure survival predictions are more closely calibrated with the observed survival of individuals diagnosed subsequently.

Drug survival of adalimumab, ustekinumab and secukinumab in patients with psoriasis: a prospective cohort study from the British Association of Dermatologists Biologics and Immunomodulators Register (BADBIR).

Real-world biologic drug survival is an important proxy measure for effectiveness. Predictors of drug survival may help patients with psoriasis choose between biologic therapies. (i) To assess the relative drug survival of adalimumab, ustekinumab and secukinumab in patients with psoriasis. (ii) To investigate predictors of biologic drug survival. A prospective cohort study was performed in the British Association of Dermatologists Biologics and Immunomodulators Register (BADBIR) between November 2007 and August 2019. We performed survival analysis and fitted a flexible parametric survival model for biologic discontinuation due to ineffectiveness. In total 9652 patients were included: 5543 starting on adalimumab (57·4%), 991 on secukinumab (10·3%) and 3118 on ustekinumab (32·3%). The overall drug survivals of adalimumab, secukinumab and ustekinumab in year 1 were 0·78 [95% confidence interval (CI) 0·77-0·79], 0·88 (95% CI 0·86-0·91) and 0·88 (95% CI 0·87-0·89), respectively. The adjusted hazard ratios (adjHRs) for discontinuation of adalimumab and secukinumab compared with ustekinumab were 2·11 (95% CI 1·76-2·54) and 0·67 (95% CI 0·40-1·11), respectively. The presence of psoriatic arthritis predicted for survival in the adalimumab and secukinumab cohorts (adjHR 0·67, 95% CI 0·51-0·88 and 0·70, 95% CI 0·40-1·24, respectively), but for discontinuation in the ustekinumab cohort (adjHR 1·42, 95% CI 1·12-1·81). Previous exposure to biologic therapies predicted for discontinuation in the ustekinumab and secukinumab cohorts (adjHR 1·54, 95% CI 1·26-1·89 and 1·49, 95% CI 0·91-2·45, respectively) and for survival in the adalimumab cohort (adjHR 0·71, 95% CI 0·55-0·92). Secukinumab and ustekinumab have similar sustained drug survival, while adalimumab has a lower drug survival in patients with psoriasis. Psoriatic arthritis and previous biologic experience were predictors with differential effects between the biologic therapies. What is already known about this topic? There is conflicting evidence over the real-world drug survival of secukinumab in patients with psoriasis. Data from registries to date suggest that secukinumab has a lower drug survival than that reported from clinical trials. What does this study add? This study found that secukinumab and ustekinumab had similar sustained drug survival in the real world, while the drug survival of adalimumab was lower, suggesting that the real-world drug survival of secukinumab is higher than previously reported. We found that psoriatic arthritis and previous biologic experience had differential effects on drug discontinuation in the three biologic cohorts. These predictors may help patients and clinicians choose the most appropriate biologic therapy.

Does minimum follow-up time post-diagnosis matter? An assessment of changing loss of life expectancy for people with cancer in Western Australia from 1982 to 2016

BackgroundCancer survival has improved in Western Australia (WA) over recent decades. Loss of life expectancy (LOLE) is a useful measure for assessing cancer survival at a population-level. Some previous studies estimating LOLE have required a minimum follow-up beyond diagnosis to reduce the impact of modelled extrapolation, while others have not. The first aim of this study was to assess the impact of minimum length of follow-up on LOLE estimates for people diagnosed in 2006 with female breast, colorectal, prostate, lung, cervical, combined oesophageal and stomach cancers, and melanoma. Based on these results, the second aim was to assess temporal changes in LOLE for these cancer types for diagnoses between 1982 and 2016. MethodsPerson-level linked cancer registry and mortality data were used for invasive primary cancer diagnoses for WA residents aged 15–89 years. The analysis for aim one included cases diagnosed from 1982 to the end of 2006, followed to the end of 2006 (i.e. no minimum follow-up), 2011 (i.e. five years minimum follow-up, assuming survival) or 2016 (i.e. 10 years minimum follow-up). To achieve the second study aim, the diagnostic period was extended to the end of 2016. Life expectancy estimates were obtained after fitting flexible parametric relative survival models. Single-year age and sex-specific death rates were used as a reference to estimate LOLE and proportionate loss of life expectancy. ResultsTemporal changes were not reported for prostate, cervical, oesophageal and stomach cancers or melanoma, due to differences in LOLE estimates by minimum follow-up time, or estimate imprecision. Marked reductions in LOLE were observed for female breast and colorectal cancer. There was minimal absolute reduction for lung cancer, where LOLE remained high. ConclusionThis study considered the appropriateness of including recent cancer diagnoses when assessing temporal changes in LOLE, finding variation in estimates with differing minimum follow-up or high parameter uncertainty for most included cancer types. Temporal changes in LOLE in-turn reflected changes in the life expectancy of the general population, cancer detection and management. These factors must be considered when estimating and interpreting LOLE estimates.

HIV programmatic outcomes following implementation of the 'Treat-All' policy in a public sector setting in Eswatini: a prospective cohort study.

IntroductionThe Treat‐All policy – antiretroviral therapy (ART) initiation irrespective of CD4 cell criteria – increases access to treatment. Many ART programmes, however, reported increasing attrition and viral failure during treatment expansion, questioning the programmatic feasibility of Treat‐All in resource‐limited settings. We aimed to describe and compare programmatic outcomes between Treat‐All and standard of care (SOC) in the public sectors of Eswatini.MethodsThis is a prospective cohort study of ≥16‐year‐old HIV‐positive patients initiated on first‐line ART under Treat‐All and SOC in 18 health facilities of the Shiselweni region, from October 2014 to March 2016. SOC followed the CD4 350 and 500 cells/mm3 treatment eligibility thresholds. Kaplan‐Meier estimates were used to describe crude programmatic outcomes. Multivariate flexible parametric survival models were built to assess associations of time from ART initiation with the composite unfavourable outcome of all‐cause attrition and viral failure.ResultsOf the 3170 patients, 1888 (59.6%) initiated ART under Treat‐All at a median CD4 cell count of 329 (IQR 168 to 488) cells/mm3 compared with 292 (IQR 161 to 430) (p < 0.001) under SOC. Although crude programme retention at 36 months tended to be lower under Treat‐All (71%) than SOC (75%) (p = 0.002), it was similar in covariate‐adjusted analysis (adjusted hazard ratio [aHR] 1.06, 95% CI 0.91 to 1.23). The hazard of viral suppression was higher for Treat‐All (aHR 1.12, 95% CI 1.01 to 1.23), while the hazard of viral failure was comparable (Treat‐All: aHR 0.89, 95% CI 0.53 to 1.49). Among patients with advanced HIV disease (n = 1080), those under Treat‐All (aHR 1.13, 95% CI 0.88 to 1.44) had a similar risk of an composite unfavourable outcome to SOC. Factors increasing the risk of the composite unfavourable outcome under both interventions were aged 16 to 24 years, being unmarried, anaemia, ART initiation on the same day as HIV care enrolment and CD4 ≤ 100 cells/mm3. Under Treat‐All only, the risk of the unfavourable outcome was higher for pregnant women, WHO III/IV clinical stage and elevated creatinine.ConclusionsCompared to SOC, Treat‐All resulted in comparable retention, improved viral suppression and comparable composite outcomes of retention without viral failure.

Temporal trends in age at menarche and age at menopause: a population study of 312 656 women in Norway.

STUDY QUESTIONHave mean age at menarche or mean age at natural menopause changed from the 1939 birth cohort to the 1964 birth cohort?SUMMARY ANSWERWe estimated a minor decrease in mean age at menarche and an increase by nearly 3 years in mean age at natural menopause.WHAT IS KNOWN ALREADYIn the Western world, age at menarche decreased across birth cohorts from the early 1800s until the 1950s. Whether mean age at menarche has continued to decrease in birth cohorts after the 1950s remains uncertain. It is also uncertain whether mean age at natural menopause has changed across birth cohorts.STUDY DESIGN, SIZE, DURATIONWe performed a retrospective population study of 312 656 women who were born in Norway during the years 1936–1964.PARTICIPANTS/MATERIALS, SETTING, METHODSThe data were obtained by two self-administered questionnaires from women who participated in the Norwegian breast cancer screening program (BreastScreen Norway) during the years 2006–2014. We used flexible parametric survival models with restricted cubic splines to estimate mean age at menarche, mean age at menopause and mean number of years between menarche and menopause according to the women’s year of birth. The women who were still having menstrual periods contributed with follow-up time until the time of data collection, and the women who had reported surgical removal of the uterus and/or both ovaries prior to natural menopause contributed with follow-up time until the time of surgery.MAIN RESULTS AND THE ROLE OF CHANCEThe mean age at menarche was 13.42 years (95% CI: 13.40–13.44 years) among women born during 1936–1939, and it was 13.24 years (95% CI: 13.22–13.25 years) among women born during 1960–1964. The mean age at natural menopause increased from 50.31 years (95% CI: 50.25–50.37 years) among women born during 1936–1939 to 52.73 years (95% CI: 52.64–52.82 years) among women born during 1960–1964. The mean number of years between menarche and menopause increased from 36.83 years (95% CI: 36.77–36.89 years) to 40.22 years (95% CI: 40.11–40.34 years).LIMITATIONS, REASONS FOR CAUTIONInformation about age at menarche and age at menopause was based on self-reports.WIDER IMPLICATIONS OF THE FINDINGSLate menopause is associated with increased risk of breast cancer but also with increased life expectancy. Thus, higher mean age at menopause may partly explain the increase in breast cancer incidence after menopause and the increase in life expectancy in recent time. Also, a longer interval between menarche and menopause could suggest that the number of years of female fecundity has increased.STUDY FUNDING/COMPETING INTEREST(S)This work was funded by the South-Eastern Norway Regional Health Authority [grant number 2016112 to M.S.G.] and by the Norwegian Cancer Society [grant number 6863294-2015 to E.K.B.]. The authors declare no conflicts of interest.