Effective utilisation of limited resources is a challenge for health care providers. Accurate and relevant information extracted from the length of stay distributions is useful for management purposes. Patient care episodes can be reconstructed from the comprehensive health registers, and in this paper we develop a Bayesian approach to analyse the length of care episode after a fractured hip. We model the large scale data with a flexible nonparametric multilayer perceptron network and with a parametric Weibull mixture model. To assess the performances of the models, we estimate expected utilities using predictive density as a utility measure. Since the model parameters cannot be directly compared, we focus on observables, and estimate the relevances of patient explanatory variables in predicting the length of stay. To demonstrate how the use of the nonparametric flexible model is advantageous for this complex health care data, we also study joint effects of variables in predictions, and visualise nonlinearities and interactions found in the data.
The use of antipsychotic agents has been associated with increased pneumonia risk, but although people with dementia are particularly susceptible to pneumonia, only one small study has assessed the risk of pneumonia in relation to the use of antipsychotic agents among people with Alzheimer disease (AD).
We investigated whether the incident use of antipsychotic agents, or specific antipsychotic agents, are related to a higher risk of hospitalization or death due to pneumonia in the Medication and Alzheimer Disease (MEDALZ) cohort. The cohort includes all individuals with AD who received a clinically verified AD diagnosis in Finland in 2005 to 2011 (N = 60,584; incident pneumonia, n = 12,225). A matched comparison cohort without AD (N = 60,584; incident pneumonia, n = 6,195) was used to compare the magnitude of risk. Results were adjusted for a propensity score derived from comorbidities, concomitant medications, and sociodemographic characteristics. Sensitivity analyses with case-crossover design were conducted.
The use of antipsychotic agents was associated with a higher risk of pneumonia (adjusted hazard ratio [HR], 2.01; 95% CI, 1.90-2.13) in the AD cohort and a somewhat higher risk in the non-AD cohort (adjusted HR, 3.43; 95% CI, 2.99-3.93). Similar results were observed with case-crossover analyses (OR, 2.02; 95% CI, 1.75-2.34 in the AD cohort and OR, 2.59; 95% CI, 1.77-3.79 in the non-AD cohort). The three most commonly used antipsychotic agents (quetiapine, risperidone, haloperidol) had similar associations with pneumonia risk.
Regardless of applied study design, treatment duration, or the choice of drug, the use of antipsychotic agents was associated with a higher risk of pneumonia. With observational data, we cannot fully rule out a shared causality between pneumonia and the use of antipsychotic agents, but the risk to benefit balance should be considered when antipsychotic agents are prescribed.
An excess cancer incidence of 20-25% has been identified among persons with diabetes, most of whom have type 2 diabetes. We aimed to describe the association between type 1 diabetes and cancer incidence.
Persons with type 1 diabetes were identified from five nationwide diabetes registers: Australia (2000-2008), Denmark (1995-2014), Finland (1972-2012), Scotland (1995-2012) and Sweden (1987-2012). Linkage to national cancer registries provided the numbers of incident cancers in people with type 1 diabetes and in the general population. We used Poisson models with adjustment for age and date of follow up to estimate hazard ratios for total and site-specific cancers.
A total of 9,149 cancers occurred among persons with type 1 diabetes in 3.9 million person-years. The median age at cancer diagnosis was 51.1 years (interquartile range 43.5-59.5). The hazard ratios (HRs) (95% CIs) associated with type 1 diabetes for all cancers combined were 1.01 (0.98, 1.04) among men and 1.07 (1.04, 1.10) among women. HRs were increased for cancer of the stomach (men, HR 1.23 [1.04, 1.46]; women, HR 1.78 [1.49, 2.13]), liver (men, HR 2.00 [1.67, 2.40]; women, HR 1.55 [1.14, 2.10]), pancreas (men, HR 1.53 [1.30, 1.79]; women, HR 1.25 [1.02,1.53]), endometrium (HR 1.42 [1.27, 1.58]) and kidney (men, HR 1.30 [1.12, 1.49]; women, HR 1.47 [1.23, 1.77]). Reduced HRs were found for cancer of the prostate (HR 0.56 [0.51, 0.61]) and breast (HR 0.90 [0.85, 0.94]). HRs declined with increasing diabetes duration.
Type 1 diabetes was associated with differences in the risk of several common cancers; the strength of these associations varied with the duration of diabetes.
Cites: Int J Cancer. 2002 Nov 20;102(3):258-6112397646
Individuals with Down syndrome (DS) have a predisposition to leukaemia and testicular cancer, but data on the incidence of cancers are yet sparse. A cohort of 3,581 persons with DS was identified from a National Registry of Finnish persons with intellectual disability collected between 1978 and 1986 and followed-up for cancer incidence until 2002. Standardised incidence ratios (SIRs) were defined as ratios of observed number of cancer cases to those expected from the national cancer incidence rates, by age and sex. The overall cancer risk was equal to that of the general population, but a significantly high risk of leukaemia (SIR 10.5, CI 95% 6.6-15.8) and testicular cancer (SIR4.8, CI 95% 1.8-10.4) was found.
Cemented hemiarthroplasty is preferred in treating displaced fractures of the femoral neck in the elderly. The cementing process may cause a fat embolism, leading to serious complications or death. In this study, we wanted to determine whether use of uncemented hemiarthroplasty (HA) would lead to reduced mortality and whether there are differences in the complications associated with these different types of arthroplasty.
From the PERFECT database, which combines information from various treatment registries, we identified 25,174 patients who were treated with hemiarthroplasty for a femoral neck fracture in the years 1999-2009. The primary outcome was mortality. Secondary outcomes were reoperations, complications, re-admissions, and treatment times.
Mortality was lower in the first postoperative days when uncemented HA was used. At 1 week, there was no significant difference in mortality (3.9% for cemented HA and 3.4% for uncemented HA; p = 0.09). This was also true after one year (26% for cemented HA and 27% for uncemented HA; p = 0.1). In patients treated with uncemented HA, there were significantly more mechanical complications (3.7% vs. 2.8%; p
Cites: Forensic Sci Int. 2003 Jan 28;131(2-3):113-2412590049
Cites: Stat Med. 2013 Aug 30;32(19):3388-41423508673
Cites: J Bone Joint Surg Am. 2005 Sep;87(9):2122-3016140828
Information on medicine use among coronary heart disease (CHD) patients with diabetes in unselected patient populations is scarce. This study examines the use of medication to prevent new cardiac events among newly diagnosed CHD patients with diabetes comparing them to patients without diabetes and examines socioeconomic differences in medicine use in these patient groups.
Data on CHD patients (43,501 men and 31,125 women) with or without diabetes were individually linked from nationwide registers (covering both patients treated in ambulatory and in hospital inpatient care). Age-standardised rates for medication use were calculated and differences between patient groups examined using Poisson regression.
beta-blocker use was high in all patient groups in 1997-2002, angiotensin-converting enzyme (ACE) inhibitor and angiotensin II antagonist use increased and remained higher among patients with diabetes. More than half of men and women with diabetes used ACE inhibitors and one out of five used angiotensin II antagonists in 2002. Lipid-lowering medication use increased, especially among women. In 1997-98 it was lower in lower socioeconomic groups; among men with diabetes the use remained lower than among others.
beta-blocker use was constant and ACE inhibitor and angiotensin II antagonist use increased. Lipid-lowering medication use increased considerably after a health insurance reform in 2000, in which elevated reimbursement of drug costs (75%) was extended to include all CHD patients with hyperlipidaemia.Socioeconomic differences in medication use disappeared after the reform. However, lipid-lowering medication use remained at a lower level among men with diabetes, suggesting that their treatment did not follow guidelines.
In many countries, psychiatric services have been reformed by reducing the size of hospitals and developing community mental-health services. We investigated this reform by assessing the relation between suicide risk and different ways of organising mental-health services.
We did a nationwide comprehensive survey of Finnish adult mental-health service units between Sept 1, 2004, and March 31, 2005. From health-care or social-care officers of 428 municipalities, we asked for information, classified according to the European service mapping schedule, about adult mental-health services. For each municipality, we measured age-adjusted and sex-adjusted suicide risk, pooled between 2000 and 2004, and then adjusted for register-derived socioeconomic factors.
A wide variety of outpatient services (relative risk [RR] 0.92, 95% CI 0.87-0.96), prominence of outpatient versus inpatient services (0.93, 0.89-0.97), and 24-h emergency services (0.84, 0.75-0.92) were associated with decreased death rates from suicide. However, after adjustment for socioeconomic factors, only the prominence of outpatient services was associated with low suicide rate (0.94, 0.90-0.98). We replicated this finding even after adjustment for organisational changes and inpatient treatment.
Well-developed community mental-health services are associated with lower suicide rates than are services oriented towards inpatient treatment provision. These data are consistent with the idea that population mental health can be improved by use of multifaceted, community-based, specialised mental-health services.
Academy of Finland.
Comment In: Lancet. 2009 Jan 10;373(9658):99-10019097637
The extent of random financial risk involved in the Finnish bed-day and Diagnosis Related Groups (DRG) based hospital pricing systems were estimated and compared using parametric and simulation methods. DRG based payment schemes were found to provide significantly better protection against financial risk for municipalities, but municipality's size was the main determinant of financial risk. Small municipalities should use longer contracts between hospitals or form bigger purchaser-organisations for risk pooling. In addition, the current risk management system proved to be ineffective in decreasing the random variation in total costs.
There are no actual validation studies of the Finnish Arthroplasty Register (FAR), and only a few studies about the accuracy of self-reported hip and knee arthroplasty exist. Therefore, we examine how reliably total hip (THA) and knee (TKA) arthroplasties can be identified from multiple data sources, including self-reports, the hospital discharge register, the arthroplasty register, and medical records.
Data from the FAR and from the Finnish Hospital Discharge Register (FHDR) during the years 1980-2010 were cross-checked to identify all THA and TKA events for the Kuopio Osteoporosis Risk Factor and Prevention Study cohort (n = 14,220). Unclear events were further checked from the medical records. After establishing a gold standard, by referring to confirmed THAs and TKAs, we examined the validity of self-reports in identifying the prevalent population with THA/TKA and in identifying incident THA/TKA.
Completeness of 2820 total arthroplasty events was 96.1% in FAR and 98.3% in FHDR. The self-reports had 95.1% sensitivity and 92.9% positive predictive value (PPV) to identify population with THA and for TKA sensitivity was 94.6% and PPV 95.2%. Self-reports' sensitivity of finding the actual surgery events was 65.3% and PPV 85.4% for THA and for TKA sensitivity was 62.9% and PPV 83.4%.
The best way to identify THAs and TKAs in Finland is to combine data from the FAR and the FHDR. Self-reports can be considered as suitable to identify the prevalent population with THA/TKA, and they do not work as well to identify the actual surgery events.
The association between diabetes and depression is well demonstrated. Less is known about the trends in use of antidepressants in the rapidly growing population of diabetics. We examined trends in antidepressant medication use during 1997-2007 in Finland among persons with or without diabetes using register-based data on both diabetes and antidepressant use.
The diabetes population was obtained from the FinDM II database including 50,027 persons with insulin treated (ITDM) and 346,290 persons with non-insulin treated diabetes (NITDM) identified from several administrative registers. Data on persons without diabetes were obtained from the yearly population statistics and their antidepressant use from the register for refunded prescription medicine costs covering all medicine purchases of non-institutionalised residents. Differences in trends and prevalence were examined using the binomial regression model.
Antidepressant use was more common among persons with diabetes in all age groups and each study year among both genders (prevalence ratios (RR) 1.4-2.2 for women and 1.7-2.2 for men). Prevalence was both higher (RR 2.0-2.2 women, 1.9-2.2 men), and increased more rapidly among younger persons with NITDM.
The use of register data linked using unique personal identifiers allowed us to identify a total cohort of persons with diabetes, to separate between ITDM and NITDM patients and to examine patterns of antidepressant use in populations with and without diabetes during an 11 year study period. Our results suggest that more attention should be focused on psychological well-being in those with diabetes and especially young people in risk of type 2 diabetes.
Comment In: Pharmacoepidemiol Drug Saf. 2011 Sep;20(9):1001-321919112
Comment In: Pharmacoepidemiol Drug Saf. 2012 Apr;21(4):45222495913