Modeling Zero-Inflated and Overdispersed Count Data With Application to Psychiatric Inpatient Service Use
| dc.contributor.advisor | Feng, Cindy | |
| dc.contributor.advisor | Balbuena, Lloyd | |
| dc.contributor.committeeMember | Pahwa, Punam | |
| dc.contributor.committeeMember | Khan, Shahed | |
| dc.contributor.committeeMember | Janzen, Bonnie | |
| dc.creator | Sharker, Sharmin | |
| dc.date.accessioned | 2019-08-02T14:02:22Z | |
| dc.date.available | 2019-08-02T14:02:22Z | |
| dc.date.created | 2019-07 | |
| dc.date.issued | 2019-08-02 | |
| dc.date.submitted | July 2019 | |
| dc.date.updated | 2019-08-02T14:02:22Z | |
| dc.description.abstract | Psychiatric disorders can be characterized as behavioral or mental states that cause significant distress and impaired personal functioning. Such disorders may occur as a single episode or persistent, relapsing, and perhaps leading to suicidal behaviours. The exact causes of psychiatric disorders are hard to determine but easy access to health care services can help to reduce the severity of the states. Inpatient psychiatric hospitalization is not only an expensive mode of treatment but also may represent the quality of health care system. The aim of this study was to investigate the factors associated with repeated hospitalizations among the patients with psychiatric illness, which may help the policy makers to target the high-risk groups in a more focused manner. The count of hospitalizations for psychiatric patients may be zero during a period of time for the huge majority of patients rather than a positive count. A common strategy to handle excessive zeros is to use zero-inflated models or hurdle models. In the field of health services research of mental health, very little literature is available comparing the relative fits of zero-inflated distributions and other count distributions to empirical data. A large linked administrative database consisting of 200,537 patients with psychiatric diagnosis in the years of 2008-2012 was used in this thesis. Various counts regression models were considered for analyzing the hospitalization rate among patients with psychiatric disorders within 3, 6 and 9 months follow-up since index visit date. The covariates for this study consist of sociodemographic and clinical characteristics of the patients. According to the Akaike Information Criteria, Vuong’s test and randomized quantile residuals, the hurdle negative binomial model was the best model. Our results showed that hospitalization rate depends on the patients’ socio-demographic characteristics and also on disease types. It also showed that having previously visited a general physician served a protective role for psychiatric hospitalization during our study period. Patients who had seen an outpatient psychiatrist were more likely to have a higher number of psychiatric hospitalizations. This may indicate that psychiatrists tend to see patients with more severe illnesses, who require hospital-based care for managing their illness. Having earlier and greater access to outpatient psychiatrist and community-based mental health care may alleviate the need for hospital-based psychiatric care. | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.uri | http://hdl.handle.net/10388/12237 | |
| dc.subject | Zero-Inflated models | |
| dc.title | Modeling Zero-Inflated and Overdispersed Count Data With Application to Psychiatric Inpatient Service Use | |
| dc.type | Thesis | |
| dc.type.material | text | |
| thesis.degree.department | School of Public Health | |
| thesis.degree.discipline | Biostatistics | |
| thesis.degree.grantor | University of Saskatchewan | |
| thesis.degree.level | Masters | |
| thesis.degree.name | Master of Science (M.Sc.) |