Identifying subtypes of chronic kidney disease with machine learning: development, internal validation and prognostic validation using linked electronic health records in 350,067 individuals.

BACKGROUND: Although chronic kidney disease (CKD) is associated with high multimorbidity, polypharmacy, morbidity and mortality, existing classification systems (mild to severe, usually based on estimated glomerular filtration rate, proteinuria or urine albumin-creatinine ratio) and risk prediction models largely ignore the complexity of CKD, its risk factors and its outcomes. Improved subtype definition could improve prediction of outcomes and inform effective interventions. METHODS: We analysed individuals ≥18 years with incident and prevalent CKD (n = 350,067 and 195,422 respectively) from a population-based electronic health record resource (2006-2020; Clinical Practice Research Datalink, CPRD). We included factors (n = 264 with 2670 derived variables), e.g. demography, history, examination, blood laboratory values and medications. Using a published framework, we identified subtypes through seven unsupervised machine learning (ML) methods (K-means, Diana, HC, Fanny, PAM, Clara, Model-based) with 66 (of 2670) variables in each dataset. We evaluated subtypes for: (i) internal validity (within dataset, across methods); (ii) prognostic validity (predictive accuracy for 5-year all-cause mortality and admissions); and (iii) medications (new and existing by British National Formulary chapter). FINDINGS: After identifying five clusters across seven approaches, we labelled CKD subtypes: 1. Early-onset, 2. Late-onset, 3. Cancer, 4. Metabolic, and 5. Cardiometabolic. Internal validity: We trained a high performing model (using XGBoost) that could predict disease subtypes with 95% accuracy for incident and prevalent CKD (Sensitivity: 0.81-0.98, F1 score:0.84-0.97). Prognostic validity: 5-year all-cause mortality, hospital admissions, and incidence of new chronic diseases differed across CKD subtypes. The 5-year risk of mortality and admissions in the overall incident CKD population were highest in cardiometabolic subtype: 43.3% (42.3-42.8%) and 29.5% (29.1-30.0%), respectively, and lowest in the early-onset subtype: 5.7% (5.5-5.9%) and 18.7% (18.4-19.1%). MEDICATIONS: Across CKD subtypes, the distribution of prescription medication classes at baseline varied, with highest medication burden in cardiometabolic and metabolic subtypes, and higher burden in prevalent than incident CKD. INTERPRETATION: In the largest CKD study using ML, to-date, we identified five distinct subtypes in individuals with incident and prevalent CKD. These subtypes have relevance to study of aetiology, therapeutics and risk prediction. FUNDING: AstraZeneca UK Ltd, Health Data Research UK.

Using national electronic health records for pandemic preparedness: validation of a parsimonious model for predicting excess deaths among those with COVID-19-a data-driven retrospective cohort study.

OBJECTIVES: To use national, pre- and post-pandemic electronic health records (EHR) to develop and validate a scenario-based model incorporating baseline mortality risk, infection rate (IR) and relative risk (RR) of death for prediction of excess deaths. DESIGN: An EHR-based, retrospective cohort study. SETTING: Linked EHR in Clinical Practice Research Datalink (CPRD); and linked EHR and COVID-19 data in England provided in NHS Digital Trusted Research Environment (TRE). PARTICIPANTS: In the development (CPRD) and validation (TRE) cohorts, we included 3.8 million and 35.1 million individuals aged ≥30 years, respectively. MAIN OUTCOME MEASURES: One-year all-cause excess deaths related to COVID-19 from March 2020 to March 2021. RESULTS: From 1 March 2020 to 1 March 2021, there were 127,020 observed excess deaths. Observed RR was 4.34% (95% CI, 4.31-4.38) and IR was 6.27% (95% CI, 6.26-6.28). In the validation cohort, predicted one-year excess deaths were 100,338 compared with the observed 127,020 deaths with a ratio of predicted to observed excess deaths of 0.79. CONCLUSIONS: We show that a simple, parsimonious model incorporating baseline mortality risk, one-year IR and RR of the pandemic can be used for scenario-based prediction of excess deaths in the early stages of a pandemic. Our analyses show that EHR could inform pandemic planning and surveillance, despite limited use in emergency preparedness to date. Although infection dynamics are important in the prediction of mortality, future models should take greater account of underlying conditions.

Significant reduction in chronic kidney disease progression with sodium-glucose cotransporter-2 inhibitors compared to dipeptidyl peptidase-4 inhibitors in adults with type 2 diabetes in a UK clinical setting: An observational outcomes study based on international guidelines for kidney disease.

AIMS: To confirm the reno-protective effects of sodium-glucose cotransporter-2 (SGLT2) inhibitors compared with dipeptidyl peptidase-4 (DPP-4) inhibitors on the onset and progression of chronic kidney disease (CKD) in routine clinical practice. MATERIALS AND METHODS: We conducted a retrospective cohort study using the Clinical Practice Research Datalink Aurum database linked to Hospital Episode Statistics. The primary outcome was risk of the composite CKD endpoint based on the recent consensus guidelines for kidney disease: >40% decline in estimated glomerular filtration rate (eGFR), kidney death or end-stage kidney disease (ESKD; a composite of kidney transplantation, maintenance of dialysis, sustained low eGFR <15 ml/min/1.73m² or diagnosis of ESKD). Secondary outcomes were components of the composite CKD endpoint, analysed separately. Patients were propensity-score-matched 1:1 for SGLT2 inhibitor versus DPP-4 inhibitor use. RESULTS: A total of 131 824 people with type 2 diabetes (T2D) were identified; 79.0% had no known history of CKD. During a median follow-up of 2.1 years, SGLT2 inhibitor initiation was associated with lower risk of progression to composite kidney endpoints than DPP-4 inhibitor initiation (7.48 vs. 11.77 events per 1000 patient-years, respectively). Compared with DPP-4 inhibitor initiation, SGLT2 inhibitor initiation was associated with reductions in the primary composite CKD endpoint (hazard ratio [HR] 0.64, 95% confidence interval [CI] 0.56-0.74), all-cause mortality (HR 0.74, 95% CI 0.64-0.86) and ESKD (HR 0.37, 95% CI 0.25-0.55), reduced the rate of sustained low eGFR (HR 0.33, 95% CI 0.19-0.57), and reduced diagnoses of ESKD in primary care (HR 0.04, 95% CI 0.01-0.18). Results were consistent across subgroup and sensitivity analyses. CONCLUSIONS: In adults with T2D, initiation of an SGLT2 inhibitor was associated with a significantly reduced risk of CKD progression and death compared with initiation of a DPP-4 inhibitor.

A retrospective cohort study predicting and validating impact of the COVID-19 pandemic in individuals with chronic kidney disease.

Chronic kidney disease (CKD) is associated with increased risk of baseline mortality and severe COVID-19, but analyses across CKD stages, and comorbidities are lacking. In prevalent and incident CKD, we investigated comorbidities, baseline risk, COVID-19 incidence, and predicted versus observed one-year excess death. In a national dataset (NHS Digital Trusted Research Environment [NHSD TRE]) for England encompassing 56 million individuals), we conducted a retrospective cohort study (March 2020 to March 2021) for prevalence of comorbidities by incident and prevalent CKD, SARS-CoV-2 infection and mortality. Baseline mortality risk, incidence and outcome of infection by comorbidities, controlling for age, sex and vaccination were assessed. Observed versus predicted one-year mortality at varying population infection rates and pandemic-related relative risks using our published model in pre-pandemic CKD cohorts (NHSD TRE and Clinical Practice Research Datalink [CPRD]) were compared. Among individuals with CKD (prevalent:1,934,585, incident:144,969), comorbidities were common (73.5% and 71.2% with one or more condition[s] in respective data sets, and 13.2% and 11.2% with three or more conditions, in prevalent and incident CKD), and associated with SARS-CoV-2 infection, particularly dialysis/transplantation (odds ratio 2.08, 95% confidence interval 2.04-2.13) and heart failure (1.73, 1.71-1.76), but not cancer (1.01, 1.01-1.04). One-year all-cause mortality varied by age, sex, multi-morbidity and CKD stage. Compared with 34,265 observed excess deaths, in the NHSD-TRE and CPRD databases respectively, we predicted 28,746 and 24,546 deaths (infection rates 10% and relative risks 3.0), and 23,754 and 20,283 deaths (observed infection rates 6.7% and relative risks 3.7). Thus, in this largest, national-level study, individuals with CKD have a high burden of comorbidities and multi-morbidity, and high risk of pre-pandemic and pandemic mortality. Hence, treatment of comorbidities, non-pharmaceutical measures, and vaccination are priorities for people with CKD and management of long-term conditions is important during and beyond the pandemic.

Determinants of referral of women with urinary incontinence to specialist services: a national cohort study using primary care data from the UK.

BACKGROUND: Female urinary incontinence is underdiagnosed and undertreated in primary care. There is little evidence on factors that determine whether women with urinary incontinence are referred to specialist services. This study aimed to investigate characteristics associated with referrals from primary to specialist secondary care for urinary incontinence. METHODS: We carried out a cohort study, using primary care data from over 600 general practices contributing to the Clinical Practice Research Datalink (CPRD) in the United Kingdom. We used multi-level logistic regression to estimate adjusted odds ratios (aOR) that reflect the impact of patient and GP practice-level characteristics on referrals to specialist services in secondary care within 30 days of a urinary incontinence diagnosis. All women aged ≥18 years newly diagnosed with urinary incontinence between 1 April 2004 and 31 March 2013 were included. One-year referral was estimated with death as competing event. RESULTS: Of the 104,466 included women (median age: 58 years), 28,476 (27.3%) were referred within 30 days. Referral rates decreased with age (aOR 0.34, 95% CI 0.31-0.37, comparing women aged ≥80 with those aged 40-49 years) and was lower among women who were severely obese (aOR 0.84, 95% CI 0.78-0.90), smokers (aOR 0.94, 95% CI 0.90-0.98), women from a minority-ethnic backgrounds (aOR 0.76, 95% CI 0.65-0.89 comparing Asian with white women), women with pelvic organ prolapse (aOR 0.77, 95% CI 0.68-0.87), and women in Scotland (aOR 0.60, 95% CI 0.46-0.78, comparing women in Scotland and England). One-year referral rate was 34.0% and the pattern of associations with patient characteristics was almost the same as for 30-day referrals. CONCLUSIONS: About one in four women with urinary incontinence were referred to specialist secondary care services within one month after a UI diagnosis and one in three within one year. Referral rates decreased with age which confirms concerns that older women with UI are less likely to receive care according to existing clinical guidelines. Referral rates were also lower in women from minority-ethnic backgrounds. These finding may reflect clinicians' beliefs about the appropriateness of referral, differences in women's preferences for treatment, or other factors leading to inequities in referral for urinary incontinence.

Geographical variation in rates of surgical treatment for female stress urinary incontinence in England: a national cohort study.

OBJECTIVE: To examine geographic variation in use of surgery for female stress urinary incontinence (SUI), mainly midurethral mesh tape insertions, in the English National Health Service (NHS). DESIGN: National cohort study. SETTING: NHS hospitals. PARTICIPANTS: 27 997 women aged 20 years or older who had a first SUI surgery in an English NHS Hospital between April 2013 and March 2016 and a diagnosis of SUI at the same time as the procedure. METHODS: Multilevel Poisson regression was used to adjust for geographic differences in age, ethnicity, prevalence of long-term illness and socioeconomic deprivation. PRIMARY OUTCOME MEASURE: Rate of surgery for SUI per 100 000 women/year at two geographic levels: Clinical Commissioning Group (CCG; n=209) and Sustainability and Transformation Partnership (STP; n=44). RESULTS: The rate of surgery for SUI was 40 procedures per 100 000 women/year. Risk-adjusted rates ranged from 20 to 106 procedures per 100 000 women/year across CCGs and 24 to 69 procedures per 100 000 women/year across the STP areas. These regional differences were only partially explained by demographic characteristics as adjustment reduced variance of surgery rates by 16% among the CCGs and 35% among the STPs. CONCLUSIONS: Substantial geographic variation exists in the use of surgery for female SUI in the English NHS, suggesting that women in some areas are more likely to be treated compared with women with the same condition in other areas. The variation reflects differences in how national guidelines are being interpreted in the context of the ongoing debate about the safety of SUI surgery.

Long-term Rate of Mesh Sling Removal Following Midurethral Mesh Sling Insertion Among Women With Stress Urinary Incontinence.

Importance: There is concern about outcomes of midurethral mesh sling insertion for women with stress urinary incontinence. However, there is little evidence on long-term outcomes. Objective: To examine long-term mesh removal and reoperation rates in women who had a midurethral mesh sling insertion for stress urinary incontinence. Design, Setting, and Participants: This population-based retrospective cohort study included 95 057 women aged 18 years or older who had a first-ever midurethral mesh sling insertion for stress urinary incontinence in the National Health Service hospitals in England between April 1, 2006, and December 31, 2015. Women were followed up until April 1, 2016. Exposures: Patient and hospital factors and retropubic or transobturator mesh sling insertions. Main Outcomes and Measures: The primary outcome was the risk of midurethral mesh sling removal (partial or total) and secondary outcomes were reoperation for stress urinary incontinence and any reoperation including mesh removal, calculated with death as competing risk. A multivariable Fine-Gray model was used to calculate subdistribution hazard ratios as estimates of relative risk. Results: The study population consisted of 95 057 women (median age, 51 years; interquartile range, 44-61 years) with first midurethral mesh sling insertion, including 60 194 with retropubic insertion and 34 863 with transobturator insertion. The median follow-up time was 5.5 years (interquartile range, 3.2-7.5 years). The rate of midurethral mesh sling removal was 1.4% (95% CI, 1.3%-1.4%) at 1 year, 2.7% (95% CI, 2.6%-2.8%) at 5 years, and 3.3% (95% CI, 3.2%-3.4%) at 9 years. Risk of removal declined with age. The 9-year removal risk after transobturator insertion (2.7% [95% CI, 2.4%-2.9%]) was lower than the risk after retropubic insertion (3.6% [95% CI, 3.5%-3.8%]; subdistribution hazard ratio, 0.72 [95% CI, 0.62-0.84]). The rate of reoperation for stress urinary incontinence was 1.3% (95% CI, 1.3%-1.4%) at 1 year, 3.5% (95% CI, 3.4%-3.6%) at 5 years, and 4.5% (95% CI, 4.3%-4.7%) at 9 years. The rate of any reoperation, including mesh removal, was 2.6% (95% CI, 2.5%-2.7%) at 1 year, 5.5% (95% CI, 5.4%-5.7%) at 5 years, and 6.9% (95% CI, 6.7%-7.1%) at 9 years. Conclusions and Relevance: Among women undergoing midurethral mesh sling insertion, the rate of mesh sling removal at 9 years was estimated as 3.3%. These findings may guide women and their surgeons when making decisions about surgical treatment of stress urinary incontinence.