Statement of Problem In the past few years, there has been increased interest in the intravenous administration of antibiotics in the home. The catalyst for this interest has been the need to contain costs and decrease surgical waiting lists in hospitals. However, the scientific evidence for cost containment for home parenteral programs is scant, conflicting and confusing. This study is the first to undertake an economic model based on a cost-benefit framework encompassing both direct and the often-ignored indirect costs. It also examines patient self-administered home parenteral antibiotic therapy (HPAT) within an economic, clinical and humanistic outcomes (ECHO) structure as a source of model inputs. The modelling approach overcomes the analytical and statistical difficulties associated with these early discharge programs due to the small and diverse nature of the populations. Methods and Procedures This study uses data from the Alternate Site Infusion Service (ASIS) at Princess Alexandra Hospital, Brisbane, Australia to populate some fields in the economic model. This service is based on the patient self-administration model and provides early hospital discharge to medically stable patients who require medium to long-term intravenous antibiotics for the treatment of infectious disease. The economic model is described as a probabilistic, second-order, Monte Carlo simulation based on cost-benefit design and constructed from realistic incremental differences in costs and benefits. The multi-dimensional probabilistic sensitivity analysis is used to account for uncertainty present in some of the model inputs. Risk analysis software known as “@Risk” Version 4.5.5 Professional is used to construct the simulation model. The cost and benefit framework, and ultimately inputs, are constructed from primary data emanating from the databases of the ASIS unit, Princess Alexandra Hospital in Brisbane for the period 2001 to 2002, secondary data based on literature reviews, and expert opinion. Societal perspective is chosen to encompass areas such as loss or gain of productivity and carer burden. Clinical outcome is investigated by examining the ASIS database regarding the treatment outcome of HPAT patients. The incidence and mortality rates of nosocomial intravascular bacteraemia (NIB) for hospital and HPAT patients are based on the literature review and included as inputs in the model. Carer burden is determined by the identified carers of ASIS patients with the completion of the mailed Caregiving Distress Scale (CDS) and the Impact of Caring Scale (ICS) instruments. A labour questionnaire was designed to collect data on paid employment, students returning to education and individuals returning to normal daily activities (including forgone leisure activities). The Australian Bureau of Statistics’ (ABS) Australian Standard Classification of Occupations (ASCO) and the ABS gender-specific average weekly wages for nine major work groups are used to estimate productivity of patients and their carers. The value of a hospital bed day was modelled to estimate the value of hospitalisation at the end of a hospital length of stay to avoid using an average bed day cost. The change in the utilisation of nursing, medical and pharmacy human resources due to HPAT is also modelled to avoid using average estimates. A sensitivity analysis is conducted on the value of a hospital bed day to measure the impact on the net benefit. Results The @Risk economic model was undertaken with 10,000 iterations to capture the variability of the net benefit. The value of a hospital bed day appears to have the greatest impact on the net benefit of ASIS with the probability of NIB and incidence of death from NIB in hospital also contributing. The correlation of CDS and the ICS demonstrates a linear relationship, and a total of 93% of carers indicate that they would repeat their participation as a carer. However, most relationships between variables are not statistically significant, or clinically unimportant. Carer burden is suggested to be low in home parenteral antibiotic patients despite the ASIS unit adopting the patient self-administered form of service delivery. However, the number of patients identifying a carer is low and, therefore, there is difficulty in the collection of a suitable sized data set. HPAT failed in 10% of home patients and was fairly comparable with other studies based on the patient self-administered model of service delivery. A total of 58% of HPAT patients (total patients = 123 and 3,939 bed days) returned to work or normal activities with an estimated productivity gain of $190,045 for the economic model of 3,964 bed days. Lost productivity for carers emanating from the survey research was estimated to be $118,121 for the economic model of 3,964 bed days. A mean cost difference of $19,584 between the hospital and home NIB, based on literature probability rates, was estimated indicating a benefit to the HPAT with regard to NIB. Overall, the @Risk model for the ASIS patients representing 3,964 bed days calculated a mean net benefit of $2,450,163 for the HPAT program. A 50% reduction in a hospital bed day value to $253 still resulted in a mean net benefit of $1,447,273 – and a zero net benefit resulted when the value of a hospital bed day reached $112. This suggests that the HPAT program, based on a patient self-administered model of service delivery, represents a viable option for healthcare delivery.
| Identifer | oai:union.ndltd.org:ADTP/279231 |
| Creators | Gail Neilson |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
Page generated in 0.0021 seconds