An integrative cross-design synthesis approach to estimate the cost of illness: An applied case to the cost of depression in Catalonia

Abstract

The mix of methods, combined with a simulation model, contributed to a reduction in data gaps and, in conditions of uncertainty, supplied more complete information on the costs of depression in Catalonia. This approach to COI should be differentiated from other COI designs to allow like-with-like comparisons. A consensus on COI typology, procedures and dissemination is needed.

Cost of borderline personality disorder in Catalonia (Spain)

“…Third, this model of cost of illness is mainly based in expert estimates. Although expert driven estimates are always required to complete missing or discordant information in health economic modelling [3], the number of assumptions needed to model the costs of BPD was exceptionally high in comparison with previous costs of illness studies of schizophrenia [42], anxiety disorder [36] and depression [38] in Catalonia and in Spain. The under-reporting of BPD was firstly described in the sub-system of primary care in Catalonia [2], and it could only be compared to the case of intellectual disabilities with comorbid mental disorders.…”

“…This is a prevalence-based cost of illness study which follows a governmental and a societal perspective and a conservative approach to estimate the minimum costs incurred by persons with BPD in mental health, drug addiction, criminal justice, social care and employment. A multilevel mixed cross-design synthesis combining qualitative and quantitative methods has been followed [3].…”

“…A multilevel cross-design synthesis approach [3] has been followed. It combined ‘top-down’ secondary analysis of the official databases relevant to mental health in Catalonia and the service use by BPD patients in four ‘bottom-up’ local surveys to complete and cross-check the information obtained from the official databases on emergency care, acute hospitalisation, specialised outpatient care and temporary sick leave.…”

Framing of scientific knowledge as a new category of health care research

“…Our own experience in both types of research clearly differentiates between knowledge on the CoI obtained from framing studies where the main source of information is elicited expert knowledge based on literature reviews [63,64], and actual CoI studies based on a cross‐design synthesis of multiple sources of information, secondary analysis of all available data bases, surveys and epidemiological studies where prior expert knowledge provides an essential, but complementary role [23,65]. The same happens to BoI studies, where at present, framing studies [66] coexist at the same level of ‘evidence’ as ecological and population‐based studies [67] because of the lack of an adequate taxonomy of these two types of research.…”

“…These different approaches can be combined in hybrid designs and include expert knowledge in the quest to produce new scientific knowledge, a combination that is absolutely essential in implementation research. The new cross‐design synthesis [23] and mixed designs studies [24] combine quantitative and qualitative techniques as well as deductive, inductive and means–end‐derived elicited knowledge to represent and understand highly complex phenomena in health system research and policy, such as cost of illness studies [23], effectiveness of complex interventions [24], case‐mix development and relative efficiency of catchment areas [7], or rapid synthesis to inform policy planning [25].…”