Karel Kostev

Treatment persistence, hypoglycaemia and clinical outcomes in type 2 diabetes patients with dipeptidyl peptidase-4 inhibitors and sulphonylureas: a primary care database analysis

To investigate therapy persistence, frequency of hypoglycaemia and macrovascular outcomes among type 2 diabetes patients with dipeptidyl peptidase‐4 (DPP‐4) inhibitors (DPP‐4) and sulphonylureas (SU).

Isopropanolic black cohosh extract and recurrence-free survival after breast cancer.

OBJECTIVE To investigate the influence of an isopropanolic Cimicifuga racemosa extract (iCR) on recurrence-free survival after breast cancer, including estrogen-dependent tumors. METHODS This pharmacoepidemiologic observational retrospective cohort study examined breast cancer patients treated at general, gynecological and internal facilities linked to a medical database in Germany. The main endpoint was disease-free survival following a diagnosis of breast cancer. The impact of treatment with iCR following diagnosis was analyzed by Cox-proportional hazards models, controlling for age and other confounders. RESULTS Of 18,861 patients, a total of 1,102 had received an iCR therapy. The mean overall observation time was 3.6 years. Results showed that iCR was not associated with an increase in the risk of recurrence but associated with prolonged disease-free survival. After 2 years following initial diagnosis, 14% of the control group had developed a recurrence, while the iCR group reached this proportion after 6.5 years. The primary Cox regression model controlling for age, tamoxifen use and other confounders demonstrated a protractive effect of iCR on the rate of recurrence (hazard ratio 0.83, 95% confidence interval 0.69 0.99). This effect remained consistent throughout all variations of the statistical model, including subgroup analyses. TNM status was unknown but did not bias the iCR treatment decision as investigated separately. Hence, it was assumed to be equally distributed between treatment groups. Correlation analyses showed good internal and external validity of the database. CONCLUSION An increase in the risk of breast cancer recurrence for women having had iCR treatment, compared to women not treated with iCR is unlikely.

Persistence and compliance of medications used in the treatment of osteoporosis - analysis using a large scale, representative, longitudinal German database

OBJECTIVE Osteoporosis can be effectively treated with a number of medications. However, high persistence and compliance are required to assure efficacy. This study analyses persistence and compliance with a variety of medical interventions including p.o., i.v. and s.c. administrations in Germany. METHODS This retrospective cohort study used a representative longitudinal database (IMS® LRx) comprising longitudinal prescription data for Germany from almost 80% of all German prescriptions of members of the German statutory health insurance system. Persistence is defined as the proportion of patients who remained on their initially prescribed therapy at 1 year. Compliance is measured indirectly based on the medication possession ratio (MPR). RESULTS A total of more than 1 million patients (1,107,482) for the period 07/2007 - 06/2009 was identified in the database who received a prescription for a bisphosphonate, strontium or PTH. Of these, 268,568 patients fulfilled further inclusion criteria and were included in the persistence and compliance analysis. At 12 months the proportion of patients that remained on treatment were 65.6% for zoledronate 5 mg; 56.6% for ibandronate i.v. 3 mg; 54.7% for PTH (teriparatide and 1-84 PTH), 51.0% for ibandronate 150 mg p.o.; 44.8% for alendronate 70 mg; 43.4% for etidronate. Other values were risedronate plus calcium 42.3%; alendronate plus vitamin D 37.8%; risedronate 35 mg 35.2%; risedronate 5 mg 30.6%; strontium ranelate 31.4% and alendronate 10 mg 17.3%. CONCLUSION Persistence and compliance during the treatment of osteoporosis were found to be insufficient. Treatment using the intravenous route and PTH showed the highest persistence and compliance rates and daily oral bisphosphonates the lowest. More effort to improve treatment compliance and persistence is needed to assure clinical efficacy.

Amputation rate and risk factors in type 2 patients with diabetic foot syndrome under real-life conditions in Germany.

AIMS To determine the risk of amputation and the influencing factors for amputation for patients with type 2 diabetes with diabetic foot syndrome. METHODS Longitudinal data from general practices in Germany (Disease Analyzer database, IMS Health) were analyzed. 3892 type 2 diabetes patients (mean age: 66.0 (SD: 10.9 years), 39.1% female) with a first-time diagnosis of diabetic foot syndrome between 01/200 and 12/2004 and at least 5 year follow-up documentation in the practices were included. The analyses of amputation-free survival were carried out using Kaplan-Meier curves and log-rank tests. A multivariate Cox regression model was fitted with the incident of diabetes-associated amputations as the dependent variable and adjusted for clinical and demographic characteristics. RESULT The cumulative incidence of diabetes-associated lower limb amputations was 18.2%. Amputations are independently associated with higher age, male gender, higher HbA1c value and longer diabetes duration but also some other diabetes complications. DISCUSSION The diabetic foot syndrome can but must not lead to a lower limb amputation. Due to the great medical and economic burden on the health system caused by diabetic complications, early therapeutic intervention is essential for patients with diabetic foot syndrome.

Are Sulfonylurea and Insulin Therapies Associated With a Larger Risk of Cancer Than Metformin Therapy? A Retrospective Database Analysis

OBJECTIVE Several meta-analyses of observational studies suggested that metformin use reduces cancer risk in type 2 diabetes. However, this result was not confirmed by the few available randomized controlled trials (RCTs), and many observational studies on metformin and cancer were potentially afflicted with time-related bias. We aimed to avoid this bias when comparing cancer incidence in users of sulfonylurea, insulin, and other diabetes medications, respectively, with cancer incidence in metformin users. RESEARCH DESIGN AND METHODS In a retrospective observational study, we used the German Disease Analyzer database with patient data from general practices throughout Germany. The study sample included 22,556 patients diagnosed with type 2 diabetes. During the median follow-up time of 4.8 years, 1,446 (6.4%) patients developed any cancer. In Cox regression analyses with either monotherapies or first diabetes medications as drug exposure, users of sulfonylurea (or insulin or other antidiabetes medications) were compared with metformin users. RESULTS In multivariable adjusted models, hazard ratios were 1.09 (95% CI 0.87–1.36) for sulfonylurea monotherapy, 1.14 (95% CI 0.85–1.55) for insulin monotherapy, and 0.94 (95% CI 0.67–1.33) for other diabetes medications compared with metformin monotherapy. Results were similar for comparison of first diabetes medications. CONCLUSIONS In a retrospective database analysis, taking into account potential time-related biases, no reduced cancer risk was found in metformin users. To clarify the association between diabetes medication and cancer risk, further well-designed observational studies and RCTs are needed.

No reduced risk of overall, colorectal, lung, breast, and prostate cancer with metformin therapy in diabetic patients: database analyses from Germany and the UK

In observational studies, a lower cancer risk was reported for patients with diabetes using metformin. However, many of these studies had shortcomings like time‐related biases. We aimed to compare the incidence rate of any cancer and some selected cancer sites in metformin, sulfonylurea, and insulin users and to reduce some major biases common in observational studies.

Risk factors for dementia diagnosis in German primary care practices.

BACKGROUND Dementia is a psychiatric condition the development of which is associated with numerous aspects of life. Our aim was to estimate dementia risk factors in German primary care patients. METHODS The case-control study included primary care patients (70-90 years) with first diagnosis of dementia (all-cause) during the index period (01/2010-12/2014) (Disease Analyzer, Germany), and controls without dementia matched (1:1) to cases on the basis of age, sex, type of health insurance, and physician. Practice visit records were used to verify that there had been 10 years of continuous follow-up prior to the index date. Multivariate logistic regression models were fitted with dementia as a dependent variable and the potential predictors. RESULTS The mean age for the 11,956 cases and the 11,956 controls was 80.4 (SD: 5.3) years. 39.0% of them were male and 1.9% had private health insurance. In the multivariate regression model, the following variables were linked to a significant extent with an increased risk of dementia: diabetes (OR: 1.17; 95% CI: 1.10-1.24), lipid metabolism (1.07; 1.00-1.14), stroke incl. TIA (1.68; 1.57-1.80), Parkinsons disease (PD) (1.89; 1.64-2.19), intracranial injury (1.30; 1.00-1.70), coronary heart disease (1.06; 1.00-1.13), mild cognitive impairment (MCI) (2.12; 1.82-2.48), mental and behavioral disorders due to alcohol use (1.96; 1.50-2.57). The use of statins (OR: 0.94; 0.90-0.99), proton-pump inhibitors (PPI) (0.93; 0.90-0.97), and antihypertensive drugs (0.96, 0.94-0.99) were associated with a decreased risk of developing dementia. CONCLUSIONS Risk factors for dementia found in this study are consistent with the literature. Nevertheless, the associations between statin, PPI and antihypertensive drug use, and decreased risk of dementia need further investigations.

Nonadherence to antiepileptic drugs in Germany A retrospective, population-based study

Objective: To assess the effect of patient and drug characteristics on medication adherence in people with epilepsy (PWE) in a large cohort representative of the German population. Methods: Information was obtained from the Disease Analyzer database that collects anonymous demographic and medical data from practice computer systems throughout Germany. From 2010 to 2013, adult PWE were retrospectively analyzed regarding demographic characteristics, comorbidities, and treatment with antiepileptic drugs (AED). Adherence was measured using the medication possession ratio (MPR). Individuals with an MPR <80% were classified as nonadherent. Multivariate regression models were applied to assess the association of different factors with adherence by calculating adjusted odds ratios (OR) with 95% confidence intervals. Results: A total of 31,317 PWE were included. The mean MPR was 81.1% (SD 25.7%) with 64.7% of patients showing good adherence (MPR >80%). Patient-related factors associated with good adherence to AED treatment were West German residence (OR 1.23, p < 0.0001) and learning disability (OR 1.80, p < 0.0001). Adherence was higher in patients treated with new than old (OR 1.52, p < 0.0001) and branded than generic AED (OR 1.44, p < 0.0001). Among the most common AED, levetiracetam achieved best adherence (OR 2.85, p < 0.0001), valproate lowest. Two or more daily dosages reduced adherence (TID vs QD: OR 0.84, p = 0.005; BID vs QD: OR 0.86, p = 0.011). Conclusions: One third of PWE treated with AED in Germany showed poor adherence, which was related to demographic characteristics and drug properties. Administration of new, well-tolerated drugs in simple dosage regimens improved AED compliance.

Predictors of hypoglycaemia in insulin-treated type 2 diabetes patients in primary care: A retrospective database analysis

AIMS To investigate the frequency and predictors (diabetes care and treatment, comorbidity) of documented hypoglycaemia in primary care patients with insulin-treated type 2 diabetes. METHODS Data from 32,545 patients (mean age: 70 (SD 11) years, 50.3% males) from 1072 practices were retrospectively analyzed (Disease Analyzer database Germany: 09/2011-08/2012). Logistic regression (≥1 documented hypoglyemia) was used to adjust for confounders (age, sex, practice characteristics, diabetes treatment regimen). RESULTS The prevalence of patients (12 months) with at least one reported hypoglycaemia was 2.2% (95% CI: 2.0-2.4%). The adjusted odds of having hypoglycemia were increased for renal failure (OR; 95% CI: 1.26; 1.16-1.37), autonomic neuropathy (1.34; 1.20-1.49), and adrenocortical insufficiency (3.08; 1.35-7.05). Patients with mental disorders including dementia (1.49; 1.31-1.69), depression (1.24; 1.13-1.35), anxiety (1.18; 1.01-1.37), and affective disorders (1.80; 1.36-2.38) also showed an increased odds of having hypoglycemia. Location of the practice in an urban area was associated with a lower odds ratio (0.74; 0.68-0.80). CONCLUSIONS Both individual patient characteristics (e.g. comorbidity) and regional factors (practice location) have a substantial impact on hypoglycaemia in primary care patients with insulin therapy.

Fracture risk in patients with newly diagnosed type 2 diabetes: a retrospective database analysis in primary care

AIMS It is unclear whether fracture risk is increased in newly diagnosed type 2 diabetes patients. In addition, fracture risk of various sites (hip, spine, upper extremity) was analysed. METHODS The study included 299,104 primary care patients from 1,072 practices who received a first type 2 diabetes diagnosis during the index period (01/2000-12/2013) (Disease Analyser, Germany). Furthermore, 299,104 non-diabetic controls were included after individual matching (1:1) to diabetes cases on age, sex, type of health insurance (private or statutory) and index date (visit at date of first diabetes diagnosis). Cumulative incidence of fractures was estimated for 10 years after index date using product-limit methods. Hazard ratios were calculated using Cox regression models adjusting for comorbidity. RESULTS Cumulative 10-year incidence of any, hip, spine, wrist/hand, forearm, and upper arm/shoulder fractures were 15.4%, 2.9%, 2.6%, 5.1%, 2.3%, and 2.3% in diabetes patients and 13.1%, 2.0%, 2.1%, 4.6%, 2.2%, and 1.7% in controls (log-rank test: all p<0.001, except wrist/hand p=0.56, forearm: p=0.54), respectively. In multivariate Cox regression models, newly diagnosed type 2 diabetes was related to an significantly increased risk of any fracture (adjusted hazard ratio, HR, 95% CI: 1.36, 1.32-1.40), as well as for hip (1.56, 1.45-1.67), spine (1.37, 1.28-1.47), wrist/hand (1.15, 1.03-1.27), forearm (1.12, 1.05-1.20), and upper arm/shoulder (1.61, 1.49-1.74) fractures. CONCLUSIONS Already few years after diabetes diagnosis, type 2 diabetes patients more frequently experienced overall, hip, spine, and upper extremity fractures, respectively. The underlying mechanisms need to be further explored in order to prevent fractures among patients with type 2 diabetes.