2007 Cohort

Dr Caitriona Cahir

Prescribing in Ireland: Assessing indicators of prescribing quality and efficiency in older populations

Institution: RCSI, 2007 cohort
Supervisors: Prof. Tom Fahey (RCSI), Prof. Kathleen Bennett (RCSI), Dr Conor Teljeur (TCD)

Abstract:        
This thesis aimed to assess indicators of prescribing appropriateness and efficiency in those aged > 70 years in Ireland by: (i) estimating the prevalence and cost of potentially inappropriate prescribing (PIP); (ii) estimating the potential cost savings, by applying clinical guidelines, for the PIP indicator with the highest prevalence rate; (iii) examining the variation in PIP between GPs; and (iv) assessing the association between PIP and adverse health outcomes.

A retrospective national population study determined the prevalence and cost of PIP in the national population aged > 70 years in Ireland in 2007 by applying a modified version of the recently developed Screening Tool of Older Person’s potentially inappropriate Prescriptions (STOPP) to the Health Services Executive Primary Care Reimbursement Services (HSE-PCRS) pharmacy claims database. The PIP indicator with the highest prevalence rate was identified and potential cost savings were estimated for a one year period by examining different scenarios in prescribing patterns according to clinical and cost-effectiveness guidelines. The variation between GPs in the prevalence of PIP and patient and GP characteristics associated with PIP were examined in a multilevel logistic regression model. A retrospective cohort study examined the association between PIP and adverse health outcomes (adverse drug events (ADEs), health related quality of life (HRQOL) and hospitalisation) in community dwelling older patients across 15 general practices.

The prevalence of PIP was 36% in those aged > 70 years in 2007. PIP had a significant impact on the national prescribing budget in 2007 (9% of overall expenditure for those aged > 70 years). The most prevalent PIP indicator was the prescription of proton pump inhibitors at maximum therapeutic dosage > 8 weeks. Regular review of patients and implementation of generic substitution and step-down therapy offered substantial cost savings.

Ninety-eight percent of GPs had at least one patient with PIP. There was evidence of significant variation in PIP between GPs but after adjustment for patient level variables this variation was no longer significant.

The STOPP criteria were significantly associated with patient reported adverse health outcomes. The expected number of ADEs and hospitals visits increased by 21% and 35% respectively after adjusting for confounders. The mean HRQOL utility was also significantly lower and patients were twice as likely to be classified as vulnerable and at risk of death, functional decline and greater health service use.

There is a high prevalence of PIP in those aged ^ 70 years in Ireland with significant cost consequences. PIP was significantly associated with patient reported adverse health outcomes. Studies of PIP are important to facilitate the design of better interventions to improve medication safety in older patients.

Visit Profile

Prof Brian Cleary

Methadone use in pregnancy: prevalence and pregnancy outcomes in an Irish obstetric cohort

Institution: TCD, 2007 cohort
Supervisors:   Prof. Deirdre Murphy (TCD), Prof. Paul Gallagher (RCSI), Prof. Tom Fahey (RCSI)

Abstract: 

Background

This thesis aimed to address current knowledge deficits relating to the issues of medication use in early pregnancy in an Irish healthcare setting and adverse perinatal outcomes in opioid-dependent pregnant women on methadone maintenance treatment.

Methods

An initial descriptive study aimed to establish population estimates of the prevalence and determinants of medication use in early pregnancy. As this study identified methadone as one of the most commonly used medications with potential for fetal harm, subsequent studies examined perinatal outcomes after the use of methadone in pregnancy. Initially, a systematic review and meta-analysis examined the impact of maternal methadone dose on neonatal abstinence syndrome (NAS). Study limitations identified in the systematic review informed the design of two cohort studies, one retrospective and one prospective, examining perinatal outcomes in opioid-dependent women on methadone maintenance treatment.

Results

The descriptive study found that four in ten women used a medication besides folic acid before booking for antenatal care and that potentially harmful medications were used by approximately 3% of women. Women were more likely to report using medications with potential for fetal harm if they had an unplanned or multiple pregnancy, or if they were unemployed, smokers, single or booked for antenatal care prior to 12 weeks’ gestation. Methadone was the most commonly reported medication with potential for fetal harm.
The systematic review and meta-analysis found small differences in the incidence of NAS in infants of women on higher compared with lower methadone doses which were statistically non-significant in analyses restricted to prospective studies or to those using an objective scoring system to diagnose NAS. Maternal methadone dose did not have a consistent statistically significant effect on the incidence of medically treated NAS.
The retrospective cohort study found that methadone exposure was associated with an increased risk of adverse perinatal outcomes even when known adverse socio-demographic factors had been accounted for. Methadone exposure was associated with an increased risk of very preterm birth (<32 weeks), being small for gestational age (birth weight <10th centile), admission to the neonatal unit and diagnosis of a major congenital anomaly. There was a significant dose-response relationship between methadone and diagnosis of NAS.
The prospective cohort study found that adverse perinatal outcomes including being small for gestational age (birth weight <10th centile) and admission to the neonatal unit were common after methadone maintenance treatment in pregnancy. The duration of hospitalisation for the neonate was affected by concomitant use of opiates, benzodiazepines and cocaine. The incidence and duration of NAS were not affected by methadone dose, though the severity of NAS, as indicated by the cumulative dose of morphine treatment for NAS, was increased with methadone doses ≥80mg.

Conclusion

Medication use in early pregnancy was common. Methadone use was associated with a range of adverse perinatal outcomes. Methadone dose did not have a consistent statistically significant effect on the incidence of NAS. Further work is required to facilitate optimal use of medicines in pregnancy and help improve outcomes for opioid-dependent pregnant women and their children.

Visit Profile

Dr Sheena McHugh

The Quality of Diabetes Care in the Community: Practice, Policy and Culture

Institution: UCC, 2007 cohort
Supervisors: Prof. Ivan Perry (UCC), Ruairi Brugha (RCSI), Colin Bradley (UCC)

Abstract:
Diabetes is an important marker of health system performance. The aim of this thesis was to investigate the ‘quality chasm’ around the organisation and delivery of diabetes care in general practice, to explore GPs’ attitudes to engaging in quality improvement and to examine efforts to improve the quality of diabetes care in Ireland from practice to policy.

A postal survey of 600 GPs was conducted to assess the organization of care. This was followed by an in-depth study using semi-structured interviews with a purposive sample of 31 GPs. The qualitative methodology was also used to examine GPs’ attitudes to engaging in quality improvement. Data were analysed using a Framework approach. A 2nd observational study assessed the quality of care in 63 practices with a special interest in diabetes. Results were benchmarked against national guidelines and standards of care in the UK. The final study, a case study of policy formulation, involved semi-structured interviews with 15 members of the Expert Advisory Group (EAG) for Diabetes. The assumptions of 3 theories of the policy process were used as analytical tools.

The survey response rate was 44% (n=262). Results suggested care delivery was largely unstructured; 45% of GPs had a diabetes register (n=157), 53% reported using guidelines (n=140), 30% had formal call recall system (n=78) and 24% had none of these organisational features (n=62). The lack of coordination between settings was a major barrier to providing optimal care leading to overburdened hospitals and avoidable duplication. The lack of remuneration for chronic disease management created costs for patients and apathy among GPs. There was also a sense of inertia around quality improvement activities at a national level, in contrast to the positive experience of local change driven by a practice ethos, leadership and ‘special interest’ in diabetes. Practices with a special interest in diabetes achieved a standard of care comparable to the UK in terms of clinical process recording and achieving clinical targets; 35% of patients reached the HbA1c target of <6.5% compared to 26% in England. With regard to diabetes policy formulation, the process was best described by the Multiple Streams Theory. The formulation of recommendations by the EAG was facilitated by agreement on the priorities although the proposed solutions were influenced by personal preferences and local capacity. In contrast national decision-making was protracted and ambiguous. The lack of impetus from senior management coupled with the lack of power conferred on the EAG impeded progress.

To date quality improvement in diabetes care has been largely dependent the ‘special interest’ of professionals. The challenge for the Irish health system is to embed this activity as part of routine practice, professional responsibility and the underlying health care culture.

Visit Profile

Dr Gerald McKenna

Comparison of tooth replacement strategies for partially dentate older patients in the Republic of Ireland: a randomised controlled clinical trial

Institution: UCC, 2007 cohort
Supervisors: Prof. Finbarr Allen (UCC), Dr Denis O’Mahony (UCC), Professor Charles Normand (TCD)

Abstract:        
This study aimed to compare two different tooth replacement strategies for partially dentate older patients; namely conventional treatment with removable partial dentures (RPD) and functionally orientated treatment based on the Shortened Dental Arch concept (SDA).

132 partially dentate patients over the age of 65 years were recruited to a randomised controlled clinical trial (RCT). Changes in Oral Health related Quality of Life were recorded using the Oral Health Impact Profile (OHIP-14). Patients completed the Mini Nutritional Assessment (MNA) and provided haematological samples. All of the costs involved in delivering and maintaining the treatment were recorded for each group.

92 patients completed the RCT (69.7%). Analysis using a mixed model analysis of Covariance (ANCOVA) showed that, compared to the RPD group, the SDA group recorded better OHIP-14 scores by an average of 1.52 points after 1 month (p=0.01) and by 1.99 points after 6 months (p=0.002). The RPD group recorded statistically significant improvements in MNA scores at 1 month (p<0.001) and 6 months (p<0.001) compared to baseline. The SDA group only recorded significant improvements in MNA scores 6 months (p=0.047) after intervention. For the haematological data recorded, the only measure which illustrated consistent statistically significant improvements after treatment intervention was Vitamin D. A cost-effectiveness analysis indicated that after 6 months, treatment for the SDA group was almost twice as cost-effective compared to the RPD group (1:1.73).

Patients in the SDA group recorded significantly better OHIP-14 scores at 1 month and 6 months compared to patients in the RPD group. The only measures which illustrated consistent significant improvements in nutritional status for either group were MNA and Vitamin D levels. Treatment based on the SDA concept was almost twice as cost-effective compared to treatment based on RPDs.

Visit Profile

Dr Catriona Murphy

The distribution and determinants of home-based social care utilisation in older people: Irish and comparative perspectives

Institution:  TCD, 2007 cohort
Supervisors:  Prof. Brendan Whelan (TCD), Prof. Charles Normand (TCD), and Prof. Hannah McGee (RCSI)
Abstract: Not available

Visit Profile

Dr Aisling Sheehan

Non-Cardiac Chest Pain: The Role of Physical, Psychosocial, and Service-Related Factors in the Persistence of Pain and Health Service Use

Institution: RCSI, 2007 cohort
Supervisors:  Prof. Hannah McGee (RCSI), Dr. Siobhan Jennings (RCSI), Dr. Brendan McAdam (Beaumont Hospital)

Abstract:       
Chest pain is one of the most frequent complaints in medical settings, yet more than half of cases have no detectable cause. Once a cardiac cause has been excluded, patients are typically discharged with a label of non-cardiac chest pain (NCCP). Patients with NCCP have been shown to have poor outcomes in terms of continued symptoms, distress, and continuing concern about heart disease. In addition, NCCP represents a significant burden to health services. The aetiology and management of NCCP is under-researched and poorly understood. This study sought to examine the predictors of persistent pain and health service use for patients in whom a cardiac diagnosis had been excluded. It also sought to explore how patients interpreted their symptoms in the context of normal test results, and the impact of their experiences with health services on these interpretations.

A mixed-methods design was adopted. A prospective cohort study was initially conducted with 145 participants with chest pain who attended exercise stress testing and had normal test results. At one-year follow-up, 69% reported continued pain. In addition, nearly half of participants had returned to their general practitioner and one in ten had attended the emergency department for the investigation of chest pain. In logistic regression analyses, the variables heartburn, pain precipitated by movement, cardiac anxiety, illness perceptions, and lack of communication about test results were predictive of persistent chest pain. When participants with continued chest pain were categorised into persistent healthcare users and non-persistent healthcare users, these variables were predominantly associated with participants with persistent health service use for chest pain. In addition, a number of psychological variables including anxiety and depression distinguished the persistent service users. Employment appeared to be a protective factor against persistent pain and related service use.

A small sample of participants from this cohort was interviewed in a qualitative study informed by the principles of Interpretative Phenomenological Analysis (IPA). Analysis revealed three predominant themes: 1) the disempowerment of normal test results; 2) limbo – the inner struggle of negating and relating to potential causes; and 3) the inadequacy of healthcare to validate and care for symptoms. The dynamic, complex process of interpreting symptoms and deciding whether to seek healthcare was illuminated.

The results indicate that interventions targeting the assessment of a potential gastro-oesophageal or musculoskeletal cause and the reduction of cardiac anxiety are likely to improve outcomes in these patients. Improved communication with patients is also indicated. Psychological factors appear to drive persistent service use and interventions targeting these are likely to reduce medical costs. Patients with NCCP are not a homogenous patient group and an individualised, stepped-care approach to management appears to be warranted.

Chest pain is one of the most frequent complaints in medical settings, yet more than half of cases have no detectable cause. Once a cardiac cause has been excluded, patients are typically discharged with a label of non-cardiac chest pain (NCCP). Patients with NCCP have been shown to have poor outcomes in terms of continued symptoms, distress, and continuing concern about heart disease. In addition, NCCP represents a significant burden to health services. The aetiology and management of NCCP is under-researched and poorly understood. This study sought to examine the predictors of persistent pain and health service use for patients in whom a cardiac diagnosis had been excluded. It also sought to explore how patients interpreted their symptoms in the context of normal test results, and the impact of their experiences with health services on these interpretations.

A mixed-methods design was adopted. A prospective cohort study was initially conducted with 145 participants with chest pain who attended exercise stress testing and had normal test results. At one-year follow-up, 69% reported continued pain. In addition, nearly half of participants had returned to their general practitioner and one in ten had attended the emergency department for the investigation of chest pain. In logistic regression analyses, the variables heartburn, pain precipitated by movement, cardiac anxiety, illness perceptions, and lack of communication about test results were predictive of persistent chest pain. When participants with continued chest pain were categorised into persistent healthcare users and non-persistent healthcare users, these variables were predominantly associated with participants with persistent health service use for chest pain. In addition, a number of psychological variables including anxiety and depression distinguished the persistent service users. Employment appeared to be a protective factor against persistent pain and related service use.

A small sample of participants from this cohort was interviewed in a qualitative study informed by the principles of Interpretative Phenomenological Analysis (IPA). Analysis revealed three predominant themes: 1) the disempowerment of normal test results; 2) limbo – the inner struggle of negating and relating to potential causes; and 3) the inadequacy of healthcare to validate and care for symptoms. The dynamic, complex process of interpreting symptoms and deciding whether to seek healthcare was illuminated.

The results indicate that interventions targeting the assessment of a potential gastro-oesophageal or musculoskeletal cause and the reduction of cardiac anxiety are likely to improve outcomes in these patients. Improved communication with patients is also indicated. Psychological factors appear to drive persistent service use and interventions targeting these are likely to reduce medical costs. Patients with NCCP are not a homogenous patient group and an individualised, stepped-care approach to management appears to be warranted.

Visit Profile