Page 69 - S. Ličen, I. Karnjuš, & M. Prosen (Eds.). (2019). Women, migrations and health: Ensuring transcultural healthcare. Koper, University of Primorska Press.
P. 69
Working on Epidemiological Data: Problems Encountered and Lessons Learned
Unfortunately, these deficiencies are generalized, to the point that the
Ministry of Health summary of the history of the introduction of the HDR
and the characteristics of this important tool ends with the paragraph ‘Lim-
itations and cautions in the use of data,’ thus recognizing the existence of
the problem. ‘It is necessary to keep in mind that the HDR database, even
if it has a very rich information content and an almost total coverage of the
Italian hospitals, has limits and criticalities in the use, such as problems of
homogeneity of compilation, completeness and accuracy for some variables
(mainly in the first years) and variation of the classification systems over the
years.’⁸
In addition to incompleteness of information, data analysis showed an
even more worrying aspect. Information about marital status and educa-
tion, in addition to being available for only about a third of hospitalizations,
revealed serious inconsistencies that are not acceptable because they un-
dermine the reliability of the database. For example, crossing education and
marital status with age, underage patients with a university degree were no-
ticed and less than ten years old girls turn out to be married and sometimes
even already separated. If there were few cases, we could think of inattention
in writing data on paper or when recording on electronic media, but dozens
of inconsistencies of this type suggest that they may also due to some kind
of systematic error.
Faced with a rather generalizes situation, according to the Ministry of
Health, it may be appropriate that the INTEGRA project guidelines will con-
tain a strong recommendation to support the completeness and correctness
of the compilation of the HDRs. The HDR database contains also fields in
which up to six different diagnoses may be recorded. Specific diagnoses are
very numerous and detailed, but they can be grouped into a smaller number
of categories, identified by the first 3 digits of the codes used for registration.
These categories may then be further grouped in 17 chapters, according to
the International Classification of Diseases ICD-9-CM. Another six fields are
reserved for the interventions to which the patient has been submitted dur-
ing the hospitalization and interventions too may be aggregated, for the
purposes of the analysis, in 17 general categories.
The plurality of recordable diagnoses and interventions caused a short-
coming detected by the analysis of the database that revealed the frequent
duplication of similar diagnoses within the same HDR card. This appears to be
⁸ See http://www.salute.gov.it/portale/temi/p2_6.jsp?id=1232&area=ricoveriOspedalieri&menu
=vuot.
67
Unfortunately, these deficiencies are generalized, to the point that the
Ministry of Health summary of the history of the introduction of the HDR
and the characteristics of this important tool ends with the paragraph ‘Lim-
itations and cautions in the use of data,’ thus recognizing the existence of
the problem. ‘It is necessary to keep in mind that the HDR database, even
if it has a very rich information content and an almost total coverage of the
Italian hospitals, has limits and criticalities in the use, such as problems of
homogeneity of compilation, completeness and accuracy for some variables
(mainly in the first years) and variation of the classification systems over the
years.’⁸
In addition to incompleteness of information, data analysis showed an
even more worrying aspect. Information about marital status and educa-
tion, in addition to being available for only about a third of hospitalizations,
revealed serious inconsistencies that are not acceptable because they un-
dermine the reliability of the database. For example, crossing education and
marital status with age, underage patients with a university degree were no-
ticed and less than ten years old girls turn out to be married and sometimes
even already separated. If there were few cases, we could think of inattention
in writing data on paper or when recording on electronic media, but dozens
of inconsistencies of this type suggest that they may also due to some kind
of systematic error.
Faced with a rather generalizes situation, according to the Ministry of
Health, it may be appropriate that the INTEGRA project guidelines will con-
tain a strong recommendation to support the completeness and correctness
of the compilation of the HDRs. The HDR database contains also fields in
which up to six different diagnoses may be recorded. Specific diagnoses are
very numerous and detailed, but they can be grouped into a smaller number
of categories, identified by the first 3 digits of the codes used for registration.
These categories may then be further grouped in 17 chapters, according to
the International Classification of Diseases ICD-9-CM. Another six fields are
reserved for the interventions to which the patient has been submitted dur-
ing the hospitalization and interventions too may be aggregated, for the
purposes of the analysis, in 17 general categories.
The plurality of recordable diagnoses and interventions caused a short-
coming detected by the analysis of the database that revealed the frequent
duplication of similar diagnoses within the same HDR card. This appears to be
⁸ See http://www.salute.gov.it/portale/temi/p2_6.jsp?id=1232&area=ricoveriOspedalieri&menu
=vuot.
67
