[Uniformity and comparability of national Health Information Systems for home care, nursing homes, and hospices in Italy]. / Uniformità e comparabilità dei flussi informativi nazionali per le cure domiciliari (flusso SIAD), l'assistenza residenziale (flusso FAR) e gli hospice (flusso HOS).

ABSTRACT

BACKGROUND: Health Information Systems (HIS) play a strategic role in the development of community healthcare services, a field still underdeveloped in Italy as shown by the COVID-19 pandemic, and their use for epidemiological purposes is increasing over the years. However, some general critical issues have been reported concerning national community health information systems, but no detailed study was found after a non-systematic review in Medline database and institutional websites. OBJECTIVES: to assess uniformity and comparability of health information collected by the national health information Systems for Home Care (SIAD), Nursing Homes (FAR), and Hospices (HOS) in Italy, three healthcare settings sharing patients with common conditions and healthcare needs. METHODS: information was gathered from current Technical Regulations Papers of the Italian Health Ministry. All variables of the three Information Systems were ordered on the basis of the characteristics provided by the ministerial documents into a single grouping model created for the purpose and their distribution compared among and between the Systems. According to the grouping model, the variables were divided in two main groups System Variables for administrative, bill, and identification-personal data and the Pathway Variables for patient's conditions and provided healthcare data. Common information content variables among all systems and between two of them were then identified, highlighting those with also identical terminology and allowed values. On the basis of the percentages of common and identical variables, uniformity in content and terminology was then calculated among all systems and between two of them. Besides, levels of content and terminology agreement have been calculated with Cohen's K matching the three Systems in all possible combinations of pairs. RESULTS: there are 70 variables in SIAD, 45 variables in FAR, and 34 in HOS. System variables are nearly the half in FAR and HOS, 22 (48.9%) and 17 (50.0%) variables, respectively. Pathway variables are prevalent only in SIAD, with 55 variables (78.6%). Only SIAD and HOS use ICD-9-CM, with 2 (2.8%) and 9 (26.5%) variables, respectively. The three Systems share 18 common variables, with other 16 common between just two of them. Considering the common variables, the total number of variables used by all the Systems is 97, with 23 System variables (23.7%) and 74 Pathway variables (76.3%). Overall, content uniformity among the three Systems is 18.5%, but becomes 60.9% considering only the System variables and 5.4% for the Pathway variables, with respectively 14 and 4 common variables. Among the common variables, 11 have the same denomination and allowed values, with an overall terminological uniformity of 11.3%. Being all of them System variables, no terminological uniformity has been found among the three Systems.The level of content agreement was fair for the couple FAR-HOS (K Cohen 0.26), but null for HOS-SIAD and SIAD-FAR (K Cohen -0.20 and <0.01). Null was also the level of terminological agreement in all the possible pairs of matched systems (K Cohen all negative). CONCLUSIONS: content and terminology uniformity of National Health Information Systems in Italy for Home Care, Nursing Homes, and Hospices has been found to be poor, with little reference to standardised classification systems and a scarce level of comparability of the information gathered in the three healthcare settings, although similarity in patients' characteristics and provided health services. Data comparability among them is mainly limited to administrative and identification-personal information, with little possibility of comparing information on patients' conditions and provided healthcare in the three settings. This scarce uniformity might undermine the contribution of national Health Information Systems in the development of community healthcare services in Italy and limit the potential of epidemiological research in this area. Also in the light of the establishment of new national Health Information Systems for Primary Care and Community Hospitals, a methodological reassessment of languages, codes, and evaluation tools used by community health services and information systems is needed.