Correlation of frailty assessment metrics in one-year follow-up of aged care residents: a sub-study of a randomised controlled trial.

INTRODUCTION: Frailty is increasingly recognised as a dynamic syndrome, with multiple causes, dimensions and consequences. There is little understanding of how those frailty assessment metrics interact over time. The aim of this study was to describe the longitudinal correlation between five frailty metrics, namely multimorbidity, muscular strength, mood alterations, cognitive capacity, and functional capacity in a cohort study of aged care (nursing home) residents. METHODS: 248 aged care residents with Frailty Index at baseline of < 0.4 and no dementia were followed for 12 months. A multimorbidity score and an activity of daily living limitation score were created using individual items of the Frailty Index. Muscular strength was measured by grip strength. Cognitive capacity was measured using the Montreal Cognitive Assessment (MoCA) test. Mood alterations were measured using the anxiety/depression screening question from EQ-5D. We analysed the inter-individual correlation at baseline, association between baseline and future change, and within-individual correlation at baseline, 6 and 12 months. RESULTS: Population analysis shows that metrics were not associated at baseline. All of the studied metrics at baseline were associated with change in 12 months, with the exception of anxiety/depression scores. Pairwise within-individual correlation was strong between MoCA and grip strength (0.13, p = 0.02) and activity of daily living (- 0.48, p < 0.001), and between activities of daily living and multimorbidity index (0.28, p < 0.001). No within-individual correlation was found between anxiety depression score and other metrics. CONCLUSION: The results suggest an interdependence between comorbidities, physical capacity, cognition and activities of daily living in aged care residents. Comprehensive measurement of frailty-related metrics may provide improved understanding of frailty progression at later life stages.

Rituximab and Pyoderma Gangrenosum: An Investigation of Disproportionality Using a Systems Biology-Informed Approach in the FAERS Database.

BACKGROUND: Studies have found an increased risk of pyoderma gangrenosum associated with rituximab. The structural properties and pharmacological action of rituximab may affect the risk of pyoderma gangrenosum. Additionally, pyoderma gangrenosum is associated with autoimmune disorders for which rituximab is indicated. OBJECTIVE: We aimed to determine whether rituximab is disproportionally associated with pyoderma gangrenosum using a systems biology-informed approach. METHODS: Adverse event reports were extracted from the US Food and Drug Administration Adverse Event Reporting System (FAERS, 2013-20). The Bayesian Confidence Propagation Neural Network Information Component was used to test for disproportionality. Comparators used to determine potential causal pathways included all other medicines, all medicines with a similar structure (monoclonal antibodies), all medicines with the same pharmacological target (CD20 antagonists) and all medicines used for the same indication(s) as rituximab. RESULTS: Thirty-two pyoderma gangrenosum cases were identified, 62.5% were female, with a median age of 48 years. There was an increased association of pyoderma gangrenosum with rituximab compared with all other medicines (exponentiated Information Component 6.75, 95% confidence interval (CI) 4.66-9.23). No association was observed when the comparator was either monoclonal antibodies or CD20 antagonists. Conditions for which an association of pyoderma gangrenosum with rituximab was observed were multiple sclerosis (6.68, 95% CI 1.63-15.15), rheumatoid arthritis (2.67, 95% CI 1.14-4.80) and non-Hodgkin's lymphoma (2.94, 95% CI 1.80-3.73). CONCLUSIONS: Pyoderma gangrenosum was reported more frequently with rituximab compared with all other medicines. The varying results when restricting medicines for the same condition suggest the potential for confounding by indication. Post-market surveillance of biologic medicines in  FAERS should consider a multi-faceted approach, particularly when the outcome of interest is associated with the underlying immune condition being treated by the medicine of interest.

Utilisation of disease modifying treatment and diversity of treatment pathways in relapsing remitting multiple sclerosis.

BACKGROUND: There is minimal information on the utilisation of Disease Modifying Treatment (DMTs) for multiple sclerosis. The appropriate and safe use of medicines is informed by utilisation studies. Outcomes can inform health interventions to improve appropriate use of medicines and post marketing surveillance activities to improve safety. OBJECTIVE: To evaluate utilisation and treatment patterns of disease modifying treatments (DMTs) for relapsing remitting multiple sclerosis (RRMS). METHODS: A representative sample of the Australian pharmaceutical benefits scheme data were analysed (2006-2016). Demographics of incident users and trends in incident and prevalent users were determined. Individual patient treatment pathways were determined by sequential initiation of medicines in two different periods (2006-2013 and 2014-2019). RESULTS: There were 20,660 patients with at least one dispensing of a DMT for RRMS during the study period (median age 41 years, 75% female). Incident and prevalent use increased by 20% and 88%, respectively. The market was responsive to 13 new listings of DMTs over the study period. Sequential treatment was found for 66% of initiators in 2006-2013 and 28.5% of initiators in 2014-2019. Diverse treatment pathways were found, with 278 and 93 unique sequences in 2006-2013 and 2014-2019, respectively. CONCLUSION: The availability of new DMTs has influenced both initial treatment choice and prevalence of users. Individualised treatment patterns and exposure to multiple medicines over time will challenge traditional pharmacovigilance systems.

Prevalence of multiple risk factors for poor outcomes associated with COVID-19 among an elderly Australian population.

BACKGROUND AND OBJECTIVES: Increasing age, male sex and various chronic conditions have been identified as important risk factors for poor outcomes from COVID-19. The aim of this study was to examine the prevalence of risk factors for poor outcomes due to COVID-19 infection in an older population. METHOD: The proportion of the population with one or more risk factors and the prevalence of individual risk factors and multiple risk factors were calculated among Department of Veterans' Affairs (DVA) clients aged ≥70 years. RESULTS: There were 103,422 DVA clients included. Of these, 79% in the community and 82% in residential aged care had at least one risk factor for poor outcomes from COVID-19. Hypertension was most prevalent, followed by chronic heart and airways disease. Over half had ≥2 risk factors, and one in five had ≥3 risk factors across multiple body systems. DISCUSSION: A substantial proportion of older Australians are at risk of poor outcomes from COVID-19 because of their multimorbid risk profile. These patients should be prioritised for proactive monitoring to avoid unintentional harm due to potential omission of care during the pandemic.

Use of medicines associated with dry mouth and dental visits in an Australian cohort.

BACKGROUND: Poor recognition of medicine-induced dry mouth can have a number of adverse effects, including difficulties with speech, chewing and swallowing dry foods, gum disease, dental caries and oral candidosis. This study examined the prevalence of use of medicines that cause dry mouth and claims for dental services funded by the Department of Veterans' Affairs (DVA) in an Australian cohort. METHODS: We used the DVA administrative health claims data to identify persons using medicines that can cause dry mouth at 1st of September 2016 and determine their DVA dental claims in the subsequent year. Results were stratified by gender, residence in community or residential aged cared facility and number of medicines. RESULTS: We identified 50 679 persons using medicines known to cause dry mouth. Of these, 72.6% were taking only one medicine that may cause dry mouth, and 21.6% were taking two. Less than half (46.2%) of all people taking at least one of these medicines had a dental claim in the following year. A smaller proportion of women (35.9%) made claims than men (56.9%), χ2  = 2248.77, P < 0.0001. CONCLUSIONS: Targeted interventions raising awareness of the relationship between some medicines and dry mouth, and the importance of dental visits are warranted.

Prevalence and changes in analgesic medication utilisation 1 year prior to total joint replacement in an older cohort of patients.

OBJECTIVES: To evaluate the prevalence and change in analgesic medications use prior to joint replacement in older patients between 2001 and 2012. METHODS: A population based epidemiological study was conducted. Opioids, non-steroidal anti-inflammatories (NSAIDs), paracetamol, corticosteroid injections, medications for neuropathic pain, hypnotics, and muscle relaxants supplied 1 year prior to total knee replacement (TKR, n = 15,517) and hip replacement (THR, n = 10,018) were assessed. Patient characteristics and surgical indication adjusted prevalence ratios (PRs) and 95% confidence intervals (CI) are provided. RESULTS: From 2001 to 2012, in the TKR cohort (median age 78.9) the prevalence of opioid use prior to surgery increased from 37% to 49% (PR = 1.01, 95% CI 1.00-1.01, P = 0.01), while in the THR cohort (median age 81.1) it increased from 44% to 54% (PR = 1.01, 95% CI 1.01-1.02, P < 0.001). Paracetamol use increased from 52% to 61% (PR = 1.0, 95% CI 1.0-1.0, P = 0.913) in the TKR cohort and from 55% to 67% (PR = 1.01, 95% CI 1.00-1.01, P = 0.005) in the THR cohort. Neuropathic pain medication use increased from 5% to 11% in the TKR cohort (PR = 1.04, 95% CI 1.02-1.06, P < 0.0001) and from 6% to 12% in the THR cohort (PR = 1.06, 95% CI 1.04-1.09, P < 0.0001). NSAID use decreased from 76% to 50% in the TKR cohort (PR = 0.96, 95% CI 0.95-0.96, P < 0.0001), and from 81% to 47% in THR cohort (PR = 0.95, 95% CI 0.94-0.95, P < 0.0001). Corticosteroid injections prevalence also decreased (TKR: 21-18%, PR = 0.97, 95% CI 0.96-0.97, P < 0.001, THR: 18-17%, PR = 0.97, 95% CI 0.96-0.98, P < 0.001). CONCLUSION: Pain medication utilization prior to joint replacement surgery changed significantly in this national older cohort of patients during the 2000s.

Patient-specific prescriber feedback can increase the rate of osteoporosis screening and treatment: results from two national interventions.

Osteoporosis interventions targeting older Australians and clinicians were conducted in 2008 and 2011 as part of a national quality improvement program underpinned by behavioural theory and stakeholder engagement. Uptake of bone mineral density (BMD) tests among targeted men and women increased after both interventions and sustained increases in osteoporosis treatment were observed among men targeted in 2008. PURPOSE: Educational interventions incorporating patient-specific prescriber feedback have improved osteoporosis screening and treatment among at-risk patients in clinical trials but have not been evaluated nationally. This study assessed uptake of BMD testing and osteoporosis medicines following two national Australian quality improvement initiatives targeting women (70-79 years) and men (75-85 years) at risk of osteoporosis. METHODS: Administrative health claims data were used to determine monthly rates of BMD testing and initiation of osteoporosis medicines in the 9-months post-intervention among targeted men and women compared to older cohorts of men and women. Log binomial regression models were used to assess differences between groups. RESULTS: In 2008 91,794 patients were targeted and 52,427 were targeted in 2011. There was a twofold increase in BMD testing after each intervention among targeted patients compared to controls (p < 0.001). Initiation of osteoporosis medicines increased by 21% among men targeted in 2008 and 34% among men targeted in 2011 compared to older controls (p < 0.01). Initiation of osteoporosis medicines among targeted women was similar to the older controls. CONCLUSION: Programs underpinned by behavioural theory and stakeholder engagement that target both primary care clinicians and patients can improve osteoporosis screening and management at the national level.

Medication use and potentially high-risk prescribing in older patients hospitalized for diabetes: a missed opportunity to improve care?

AIMS: To examine the appropriateness of medicine use and potentially high-risk prescribing before and after hospitalization for diabetes. METHODS: A retrospective cohort study of patients hospitalized for diabetes was conducted using administrative data from the Australian Government Department of Veterans' Affairs for the period between 1 January 2012 and 31 December 2012. The appropriateness of medicine use and potentially high-risk prescribing, including hyper-polypharmacy and associated treatment conflicts, were examined for the 120-day periods before and after hospitalization. RESULTS: A total of 876 patients were hospitalized for a diabetes-related complication. Of these, 25% were not dispensed an antidiabetic medicine 4 months before hospitalization and 25% had not had their HbA1c levels measured in the preceding 6 months. The use of antidiabetic medicines increased to 85% after hospitalization, with a 25.6% relative increase (95% CI 10.9-42.1) in the proportion of those dispensed insulin. The prevalence of high-risk prescribing before hospital admission was high; 70% had > 10 medicines dispensed, a third had at least one treatment conflict and half were dispensed a potentially inappropriate medicine. The use of long-acting sulphonylureas and corticosteroids had relative decreases of 46.0% (95% CI 17.0-64.9) and 29.9% (95% CI 8.8-46.0), respectively. Few changes in other high-risk prescribing patterns were observed after discharge. CONCLUSIONS: This study has identified poor medication-related care and, in particular, high-risk-prescribing in people subsequently hospitalized for diabetes. While diabetes medicine use improved after hospitalization, there was little change in potentially inappropriate medicine use, which suggests that an opportunity to improve medication use in this older vulnerable population has been missed.

Effect of a general practitioner management plan on health outcomes and hospitalisations in older patients with diabetes.

BACKGROUND: Little is known about the impact of a general practitioner management plan (GPMP) on health outcomes of patients with diabetes. AIM: To examine the impact of a GPMP on the risk of hospitalisation for diabetes. METHODS: A retrospective study using administrative data from the Australian Government Department of Veterans' Affairs was conducted (1 July 2006 to 30 June 2014) of diabetes patients either exposed or unexposed to a GPMP. The primary end-point was the risk of first hospitalisation for a diabetes-related complication and was assessed using Cox proportional hazard regression models with death as a competing risk. Secondary end-points included rates of receiving guideline care for diabetes, with differences assessed using Poisson regression analyses. RESULTS: A total of 16 214 patients with diabetes were included; 8091 had a GPMP, and 8123 did not. After 1 year, 545 (6.7%) patients with a GPMP and 634 (7.8%) of patients without a GPMP were hospitalised for a diabetes complication. There was a 22% reduction in the risk of being hospitalised for a diabetes complication (adjusted hazard ratio (HR) 0.78, 95% confidence interval (CI) 0.69-0.87, P < 0.0001) for those who received a GPMP by comparison to those who did not. Increased rates of diabetes guideline care, HbA1c claims (adjusted HR 1.29, 95% CI 1.25-1.33) and microalbuminura claims (adjusted HR 1.65, 95% CI 1.58-1.72) were observed after a GPMP. CONCLUSION: Provision of a GPMP in older patients with diabetes resulted in improved health outcomes, delaying the risk of hospitalisation at 12 months for diabetes complications. GPMP should be included as part of routine primary care for older patients with diabetes.

Increased risk of hospital admission for dehydration or heat-related illness after initiation of medicines: a sequence symmetry analysis.

WHAT IS KNOWN AND OBJECTIVE: Although several studies have identified factors which increase the risk of heat-related illness, few have assessed the contribution of medicines. To address this knowledge gap, our study aimed to assess the risk of hospital admission for dehydration or other heat-related illness following initiation of medicines. METHODS: We conducted a retrospective analysis using prescription event symmetry analysis (PESA) of 6700 veterans with incident hospital admission for dehydration or heat-related illness (ICD-10-AM codes E86, X30, T67), between 1 January 2001 and 30 June 2013. The main outcome measure was first ever hospital admission for dehydration or heat-related illness following initiation of commonly used medicines. RESULTS AND DISCUSSION: A significantly higher risk of incident hospital admission for dehydration or heat-related illness was observed following initiation of anticoagulants, cardiovascular medicines, NSAIDs, antipsychotics, antidepressants and anticholinergic agents. The risk of hospital admission for dehydration or heat-related illness ranged from 1·17 (SSRIs) to 2·79 (ACEI plus diuretic combination product). No significant association was observed between initiation of anticonvulsants, anti-Parkinson's agents, hypnotics, anxiolytics or antihistamines and hospital admission for dehydration or heat-related illness. WHAT IS NEW AND CONCLUSION: Many commonly used medicines were found to be associated with increased risk of hospitalization for dehydration or heat-related illness. Initiation of ACE inhibitors in combination with diuretics had the highest risk. Prescribers and patients should be aware of the potential for medicines to be associated with increased risk of dehydration and heat-related illness.

Construct Validity of Four Frailty Measures in an Older Australian Population: A Rasch Analysis.

Individuals identified as frail have been shown to be at an increased risk of adverse health outcomes. However, there is no gold standard frailty measure and frailty status can vary depending on the measure used, suggesting the measures perform differently. Construct validity can be used to assess a measure's performance. This study aimed to examine the construct validity of four frailty measures in an Australian older population using Rasch analysis. Frailty status among the 2087 participants aged 65 years and above from the Australian Longitudinal Study of Ageing (ALSA) was assessed using: frailty phenotype--FP, simplified frailty phenotype--SFP, frailty index--FI, and prognostic frailty score--PFS. Rasch analysis was used to assess the unidimensionality of the measures, which is the extent to which the underlying characteristic of frailty is assessed. The criteria for unidimensionality from principal component analysis of the residuals was when 50% or more of the raw variance was explained by the measures, and less than 5% was unexplained variance. Only FI meet the unidimensionality criteria with 74% of explained variance and 2.1% of unexplained variance. SFP did not show a unidimensional construct with 13.3% of explained variance and 47.1% of unexplained variance. FP and PFS had 39.6%, 18.1% and 46.5%, 8.7% of explained and unexplained variance, respectively. Our findings showed that FI has better construct validity than the other three measures in assessing frailty among the Australian older population.

Evaluation of three co-morbidity measures to predict mortality in patients undergoing total joint arthroplasty.

OBJECTIVE: To evaluate the 90 days and 1 year mortality predictive ability of the RxRisk-V, Charlson, and Elixhauser co-morbidity measures in total hip arthroplasty (THA) and total knee arthroplasty (TKA) patients. METHOD: A retrospective study of 11,848 THAs and 18,972 TKAs (2001-2002) was conducted. Death within 90 days and 1 year of the surgery were the main endpoints. Co-morbidity measures were calculated using either medication or hospitalisation history. Logistic regression models were employed and discrimination and calibration were assessed. Specifically, models with unweighted and weighted measure scores, models with the specific conditions, and a model combining conditions identified by all measures were assessed. RESULTS: In THAs, the best performing prediction models included co-morbidities from all three measures (90 days: c = 0.84, P = 0.284, 1 year: c = 0.79, P = 0.158). Individually, the model with Charlson conditions performed best at 90 days mortality (c = 0.80, P = 0.777) and the Charlson and Elixhauser performed similarly at 1 year (both c = 0.77, P > 0.05). In TKAs, the best performing prediction model included co-morbidities from all measures (90 days: c = 0.82, P = 0.349, 1 year: c = 0.78, P = 0.873). Individually, the model with Elixhauser conditions performed best with 90 days mortality (c = 0.79, P = 0.435) and all performed similarly at 1 year (c = 0.74-0.75, all P > 0.05). CONCLUSIONS: A combined model with co-morbidities identified by the Elixhauser, Charlson, and RxRisk-V was the best mortality prediction model. The RxRisk-V did not perform as well as the others. Because of the Elixhauser and Charlson's similar performance we suggest basing the choice of measurement use on factors such as the need of specific conditions and modelling limitations.

Hospitalization for drug-induced hepatotoxicity: linking Y-codes with pharmaceutical claims data to identify implicated medicines.

WHAT IS KNOWN AND OBJECTIVE: Hospital admissions associated with an adverse drug reaction are often coded to the International Classification of Diseases external cause Y-codes, denoting the medicine class deemed to cause the adverse drug reaction. Matching hospital data with outpatient dispensing data has the potential to identify the specific causative medicines but the ability to identify the causative medicines in this way has not been previously assessed. This study aimed to determine the proportion of Y-coded hospitalizations for drug-induced hepatotoxicity that could be matched with a potential causative medicine from outpatient dispensing data. METHODS: A retrospective cohort study was undertaken from 1 Jan 2005 to 30 June 2012 using data from the Australian Government Department of Veterans' Affairs of all admissions coded to drug-induced hepatotoxicity. Medicine use in the 6 months prior to hospitalization was examined to identify the probable causative medicines. RESULTS AND DISCUSSION: Thirty five admissions were identified for 31 patients. All admissions were preceded by use of medicines known to cause hepatotoxicity. Twenty four admissions had a Y-code recorded, of which 19 admissions had at least one Y-code specifying the causative medicine class (22 Y-codes). Of the 22 Y-codes, 95% could be successfully matched with a medicine from the same class that had been dispensed in the 6 months prior to admission. Further, 92% were preceded by use of multiple hepatotoxic medicines. WHAT IS NEW AND CONCLUSION: Results of our study demonstrate that hospital administrative data can be linked to prescription dispensing data to identify specific medicines suspected of causing the adverse drug reaction.

General practitioner management plans delaying time to next potentially preventable hospitalisation for patients with heart failure.

BACKGROUND: Several studies have shown that the Australian Medicare-funded chronic disease management programme can lead to improvements in care processes. No study has examined the impact on long-term health outcomes. AIMS: This retrospective cohort study assessed the association between provision of a general practitioner management plan and time to next potentially preventable hospitalisation for older patients with heart failure. METHODS: We used the Australian Government Department of Veterans' Affairs (DVA) claims database and compared patients exposed to a general practitioner management plan with those who did not receive the service. Kaplan-Meier analysis and Cox proportional hazards models were used to compare time until next potentially preventable hospitalisation for heart failure between the exposed and unexposed groups. RESULTS: There were 1993 patients exposed to a general practitioner management plan and 3986 unexposed patients. Adjusted results showed a 23% reduction in the rate of potentially preventable hospitalisation for heart failure at any time (adjusted hazard ratio, 0.77; 95% confidence interval, 0.64 to 0.92; P = 0.0051) among those with a general practitioner management plan compared with the unexposed patients. Within one year, 8.6% of the exposed group compared with 10.7% of the unexposed group had a potentially preventable hospitalisation for heart failure. CONCLUSIONS: A general practitioner management plan is associated with delayed time to next potentially preventable hospitalisation for heart failure.

Initiation and duration of proton pump inhibitors in the Australian veteran population.

AIM: To determine the duration of initial treatment with high strength proton pump inhibitors prescribed for gastro-oesophageal reflux disease in the veteran population in Australia and variance by the medical professional who initiated treatment. METHOD: Retrospective cohort study in the Australian veteran population using Department of Veterans' Affairs pharmacy claims data. Veterans who had been dispensed at least one prescription for the high strength proton pump inhibitor indicative of gastro-oesophageal reflux disease between 1 July 2004 and 30 June 2007, were eligible for full health services, and who had not been dispensed any proton pump inhibitor in the previous 12 months were included in the study. The study end-point was time to discontinuation of initial high strength proton pump inhibitor (cessation or switch to maintenance strength) stratified by the type of initial prescriber. RESULTS: Of the new users of high strength proton pump inhibitors (n= 41 041), 32% discontinued within 8 weeks, and 62% discontinued within 12 months. The median treatment duration was: 195 days (95% CI, 186-205) for patients with hospital-initiated therapy (n= 12 294), 124 days (95% CI, 121-127) for patients with treatment initiated by general practitioners (n= 25 327) and 112 days (95% CI, 104-121) for patients with treatment initiated by specialist (n= 3420). CONCLUSIONS: Only one-third of the Australian veteran patients who initiated high strength proton pump inhibitor treatment for gastro-oesophageal reflux disease discontinued (ceased or stepped down) within 8 weeks consistent with guideline recommendations. The majority continued treatment beyond recommended durations. This is not directly attributable to the initiating prescriber.

Determinants of patient-reported medication errors: a comparison among seven countries.

OBJECTIVE: Medication errors are a frequent cause of adverse drug events and a major concern for patient safety. This study compared the predictors of error among seven countries (Australia, Canada, New Zealand, the United Kingdom, the United States, Germany and the Netherlands). METHODS: We conducted a cross-sectional study using the 2007 Commonwealth Fund International Health Policy Survey data. The outcome was patient-reported error in the past 2 years. Possible predictors were studied using logistic regression. RESULTS: Eleven thousand nine hundred and ten respondents were included in this analysis, of which 1291 respondents (11%) had experienced error. Poor coordination of care was a shared concern of all seven countries [adjusted odds ratios (ORs) ranged from 2.1 (95% CI: 1.3-3.5) to 3.0 (95% CI: 2.1-4.5)]. Cost-related barriers to medical services/medicines was also a predictor in six countries [ORs ranged from 1.9 (95% CI: 1.5-2.6) to 2.6 (95% CI: 1.5-4.6)]. Other common risk factors across countries included seeing multiple specialists, multiple chronic conditions, hospitalisation and multiple emergency room visits. Cross-country heterogeneity in contributing factors included age and specific chronic condition. Number of medications, number of doctor visits, household income and education level were not associated with error in most countries. CONCLUSION: Poor coordination of care is a key risk factor in all seven countries. Cost-related barriers were also associated with an increased likelihood of error. The major challenge for all countries for error prevention is better communication among multiple healthcare providers and more structured organisation of care across healthcare settings.

Continuity of care: when do patients visit community healthcare providers after leaving hospital?

BACKGROUND/AIMS: Enhanced communication and transfer of information between healthcare providers and healthcare settings can reduce medication and healthcare errors post-hospital discharge. The timeframes within which patients access community healthcare providers post-hospital discharge are not well studied. This study aimed to determine length of time from hospital discharge until a general practice, pharmacy or specialist visit, or care planning service. METHODS: We conducted a retrospective analysis of Department of Veterans' Affairs health claims data. All 109 860 veterans hospitalized in 2006 were included. Main outcome measures were time from first hospital discharge to first claim for a general practice, pharmacy, specialist visit and/or care planning service. RESULTS: Within 30 days of hospital discharge 71% of subjects visited a general practitioner (GP), 86% had medicines dispensed from a community pharmacy and 44% saw a specialist. Median time to first pharmacy visit was 6 days (interquartile range 2-14) and 12 days for a GP visit (interquartile range 4-31). Less than 2% of the cohort received a discharge plan, case conference or medication review in the month after discharge. CONCLUSIONS: With 25% of patients having a claim for a GP service within 4 days of discharge, discharge summaries need to reach community-based health professionals within this time. Most patients visited their community pharmacy within 2 weeks of hospital discharge and before they saw their GP. Pharmacists are not routinely advised of hospitalization or provided with discharge summaries. More active engagement of this professional group in the continuum of care might improve care after hospital discharge.

Collaborative home medicines review delays time to next hospitalization for warfarin associated bleeding in Australian war veterans.

WHAT IS KNOWN AND BACKGROUND: Unintended bleeds are a common complication of warfarin therapy. We aimed to determine the impact of general practitioner-pharmacist collaborative medication reviews in the practice setting on hospitalization-associated bleeds in patients on warfarin. METHOD: We undertook a retrospective cohort study using administrative claims data for the ambulatory veteran and war widow population, Australia. Participants were veterans, war widows and their dependents aged 65 years and over dispensed warfarin. The exposed groups were those exposed to a general practitioner (GP)-pharmacist collaborative home medication review. The service includes GP referral, a home visit by an accredited pharmacist to identify medication-related problems, a pharmacist report with follow-up undertaken by the GP. The outcome measure was time to next hospitalization for bleeding. RESULTS: There were 816 veterans exposed to a home medicines review and 16,320 unexposed patients, with an average age of 81.5 years, and six to seven co-morbidities. Adjusted results showed a 79% reduction in likelihood of hospitalization for bleeding between 2 and 6 months (HR, 0.21 95% CI, 0.05-0.87) amongst those who had received a home medicines reviewed compared to the unexposed patients. No effect was seen in the time period from review to 2 months, nor in the time period 6 to 12 months post a review. WHAT IS NEW AND CONCLUSION: Medicines review in the practice setting delays time to next hospitalization for bleeding in those treated with warfarin in the period 2 to 6 months after the review, but is not sustained over time. Six monthly medication reviews may be required for patients on warfarin who are considered at high risk of bleeding.