Defining rural in Aotearoa New Zealand: a novel geographic classification for health purposes.

AIM: Describe the first specifically designed and validated five-level rurality classification for health purposes in New Zealand that is both data-driven and incorporates heuristic understandings of rurality. METHOD: Our approach involved: (1) defining the purpose and parameters of a proposed five-level Geographic Classification for Health (GCH); (2) developing a quantitative framework; (3) undertaking co-design with the National Rural Health Advisory Group (NRHAG), and extensive consultation with key stakeholders; (4) testing the validity of the five-level GCH and comparing it to previous Statistics New Zealand (Stats NZ) rurality classifications; and (5) describing rural populations and identifying differences in all-cause mortality using the GCH and previous Stats NZ rurality classifications. RESULTS: The GCH is a technically robust and heuristically valid rurality classification for health purposes. It identifies a rural population that is different to the population defined by generic Stats NZ classifications. When applied to New Zealand's Mortality Collection, the GCH estimates a rural mortality rate 21% higher than for residents of urban areas. These rural-urban disparities are masked by the generic Stats NZ classifications. CONCLUSION: The development of the five-level GCH embraces both the technical and heuristic aspects of rurality. The GCH offers the opportunity to develop a body of New Zealand rural health literature founded on a robust conceptualisation of rurality.

Developing the geographic classification for health, a rural-urban classification for New Zealand health research and policy: A research protocol.

INTRODUCTION: Rural-urban health inequities, exacerbated by deprivation and ethnicity, have been clearly described in the international literature. To date, the same inequities have not been as clearly demonstrated in Aotearoa New Zealand despite the lower socioeconomic status and higher proportion of Maori living in rural towns. This is ascribed by many health practitioners, academics and other informed stakeholders to be the result of the definitions of 'rural' used to produce statistics. AIMS: To outline a protocol to produce a 'fit-for-health purpose' rural-urban classification for analysing national health data. The classification will be designed to determine the magnitude of health inequities that have been obscured by use of inappropriate rural-urban taxonomies. METHODS: This protocol paper outlines our proposed mixed-methods approach to developing a novel Geographic Classification for Health. In phase 1, an agreed set of community attributes will be used to modify the new Statistics New Zealand Urban Accessibility Classification into a more appropriate classification of rurality for health contexts. The Geographic Classification for Health will then be further developed in an iterative process with stakeholders including rural health researchers and members of the National Rural Health Advisory Group, who have a comprehensive 'on the ground' understanding of Aotearoa New Zealand's rural communities and their attendant health services. This protocol also proposes validating the Geographic Classification for Health using general practice enrolment data. In phase 2, the resulting Geographic Classification for Health will be applied to routinely collected data from the Ministry of Health. This will enable current levels of rural-urban inequity in health service access and outcomes to be accurately assessed and give an indication of the extent to which older classifications were masking inequities.

Testing Suicide Risk Prediction Algorithms Using Phone Measurements With Patients in Acute Mental Health Settings: Feasibility Study.

BACKGROUND: Digital phenotyping and machine learning are currently being used to augment or even replace traditional analytic procedures in many domains, including health care. Given the heavy reliance on smartphones and mobile devices around the world, this readily available source of data is an important and highly underutilized source that has the potential to improve mental health risk prediction and prevention and advance mental health globally. OBJECTIVE: This study aimed to apply machine learning in an acute mental health setting for suicide risk prediction. This study uses a nascent approach, adding to existing knowledge by using data collected through a smartphone in place of clinical data, which have typically been collected from health care records. METHODS: We created a smartphone app called Strength Within Me, which was linked to Fitbit, Apple Health kit, and Facebook, to collect salient clinical information such as sleep behavior and mood, step frequency and count, and engagement patterns with the phone from a cohort of inpatients with acute mental health (n=66). In addition, clinical research interviews were used to assess mood, sleep, and suicide risk. Multiple machine learning algorithms were tested to determine the best fit. RESULTS: K-nearest neighbors (KNN; k=2) with uniform weighting and the Euclidean distance metric emerged as the most promising algorithm, with 68% mean accuracy (averaged over 10,000 simulations of splitting the training and testing data via 10-fold cross-validation) and an average area under the curve of 0.65. We applied a combined 5×2 F test to test the model performance of KNN against the baseline classifier that guesses training majority, random forest, support vector machine and logistic regression, and achieved F statistics of 10.7 (P=.009) and 17.6 (P=.003) for training majority and random forest, respectively, rejecting the null of performance being the same. Therefore, we have taken the first steps in prototyping a system that could continuously and accurately assess the risk of suicide via mobile devices. CONCLUSIONS: Predicting for suicidality is an underaddressed area of research to which this paper makes a useful contribution. This is part of the first generation of studies to suggest that it is feasible to utilize smartphone-generated user input and passive sensor data to generate a risk algorithm among inpatients at suicide risk. The model reveals fair concordance between phone-derived and research-generated clinical data, and with iterative development, it has the potential for accurate discriminant risk prediction. However, although full automation and independence of clinical judgment or input would be a worthy development for those individuals who are less likely to access specialist mental health services, and for providing a timely response in a crisis situation, the ethical and legal implications of such advances in the field of psychiatry need to be acknowledged.

'Poorly defined': unknown unknowns in New Zealand Rural Health.

There is a considerable mismatch between the population that accesses rural healthcare in New Zealand and the population defined as 'rural' using the current statistics New Zealand rural and urban categorisations. Statistics New Zealand definitions (based on population size or density) do not accurately identify the population of New Zealanders who actually access rural health services. In fact, around 40% of people who access rural health services are classified as 'urban' under the Statistics New Zealand definition, while a further 20% of people who are currently classified as 'rural' actually have ready access to urban health services. Although there is some recognition that current definitions are suboptimal, the extent of the uncertainty arising from these definitions is not widely appreciated. This mismatch is sufficient to potentially undermine the validity of both nationally-collated statistics and also any research undertaken using Statistics New Zealand data. Under these circumstances it is not surprising that the differences between rural and urban health care found in other countries with similar health services have been difficult to demonstrate in New Zealand. This article explains the extent of this mismatch and suggests how definitions of rural might be improved to allow a better understanding of New Zealand rural health.

The price of 'free'. Quantifying the costs incurred by rural residents attending publically funded outpatient clinics in rural and base hospitals.

INTRODUCTION Rural living is associated with increased costs in many areas, including health care. However, there is very little local data to quantify these costs, and their unknown quantity means that costs are not always taken into account in health service planning and delivery. AIM The aim of this study was to calculate the average time and travel costs of attending rural and base hospital outpatient clinics for rural Central Otago residents. METHODS A survey of 51 people attending rural hospital outpatient clinics. Individual costs in terms of travel and time were quantified and an average cost of both rural and base hospital attendance was calculated. RESULTS The average travel and lost time cost of attending a rural outpatient clinic was NZ$182 and 61% of respondents reported this cost had a significant effect on their weekly budget. The average cost incurred by residents associated with a base hospital attendance in Dunedin was NZ$732. DISCUSSION This study data show that costs are substantial and probably higher than most people might expect for both rural and base hospital attendances. It seems likely that these costs are a potential barrier to service access. However, the full implications of the personal costs incurred by rural residents in accessing health services are largely unstudied and therefore remain unknown in New Zealand.

Audit of a collaborative care model suggests patients with acute myocardial infarction are not disadvantaged by treatment in a rural hospital.

AIM: Recent data suggest that improvements in survival post myocardial infarction in urban hospitals have not been matched by rural hospitals. We performed an audit of a collaborative model of care between a rural hospital and its base hospital to see if this was also the case in a rural area of New Zealand, and attempted to identify reasons underlying any discrepancy, should it exist. METHODS: The medical records of all patients with a discharge diagnosis of acute myocardial infarction (AMI) over a five-year period were reviewed and data on management, medication and outcomes collected. RESULTS: 196 patients with confirmed AMI were treated from 1995-1999. There was a documented consultation with a cardiologist in 54% of cases. Seventy one per cent of patients were managed and discharged from the rural hospital, while 23% were subsequently transferred to base hospital. The in-hospital, 30-day and one-year cardiovascular mortality rates were 5.6%, 7.7% and 13.5% respectively. CONCLUSIONS: These figures compare favourably with those previously published in NZ and internationally, suggesting that rural hospitals can achieve similar outcomes to larger centres when working in close collaboration with base hospital specialists.