Skip to main content

Table 3 Overview of the model assumptions

From: Expected lifetime numbers and costs of fractures in postmenopausal women with and without osteoporosis in Germany: a discrete event simulation model

Assumptions regarding event probabilities Impact§
We applied osteoporosis prevalence rates and BMD-values from US-NHANES III reference data ***
We estimated “total” fracture probabilities by dividing fracture probabilities based on hospital cases with age-independent hospital probabilities **
We assumed highest fracture related NH probability when more than one fractures occurs in the same time interval **
We modeled fracture related entry in a NH only after a hospital stay **
We assumed that only NH entries within 3 months after a fracture may be attributable to the fracture event itself **
We applied age-dependent relative fracture risk by one standard deviation decrease in BMD to hip fractures and age-independent relative risks to other fractures **
We assumed that osteoporosis risk attributions were calculated exclusively on BMD values measured at the femoral neck **
We assumed that osteoporosis prevalence rates do not differ between women living in a NH and women who do not *
We applied relative fracture risk and prevalence for previous fractures from an international meta-analysis *
We applied relative fracture risk by one standard deviation decrease in BMD from international studies *
We applied fracture mortality data from a Canadian study *
We assumed the highest fracture excess mortality when more than one fracture occurs in the same time interval *
We allowed first entry in a NH firstly for women aged 65 or older *
We assumed that individuals in a NH remain there for their remaining lifetime *
We assumed that patients with osteoporosis will have osteoporosis for their remaining lifetime *
We allowed a maximum possible age of 100 years *
Assumptions regarding costing  
We assumed that rehabilitation probabilities after a hospital stay do not differ between women living in NH and those who do not. **
We applied Austria data for average hours of informal and professional home care by fracture type, also we assumed that the consumed hours are equivalent for hip, other femur and pelvis **
We assumed age-dependent fracture unit costs *
We assumed that the outpatient costs for humerus and wrist as well as the costs for pelvis, other femur and hip fractures are equivalent *
We took outpatient resource use data from a study considering fracture patients with inflammatory bowel disease *
We assumed that average informal and professional home care costs are only applicable for individuals not living in NH aged older than 65 years *
  1. §Expected impact on modeling results: * = low impact, ** = medium impact, *** = high impact.