ASH older people - single map

ASH older people - double map

Ambulatory sensitive hospitalisations (ASH) are mostly acute admissions that are considered potentially reducible through prophylactic or therapeutic interventions deliverable in a primary care setting[1].

In older adults in New Zealand, ASH accounts for approximately a quarter of all acute or arranged medical and surgical discharges in a year. However, determining the reasons for high or low ASH rates is complex, as it is in part a whole-of-system measure. The Agency for Healthcare Research and Quality suggests that admission rates can serve as proxy markers for primary care access and quality, with high admission rates indicating difficulty in accessing care in a timely fashion, poor care coordination or care continuity, or structural constraints such as limited supply of primary care workers[2]. This indicator can also highlight variation between different population groups that will assist with district health board (DHB) planning to reduce disparities.

Data for this Atlas were drawn from the National Minimum Dataset and present overall ASH rates by DHB. Data for conditions contributing the most to ASH rates in older adults (75–84 years) are presented. These conditions were selected with the assistance of an expert advisory group as being conditions possibly amenable to primary care interventions. The conditions were: angina/chest pain, congestive heart failure and bacterial/non-viral pneumonia.

There is considerable debate about the appropriateness of the term ASH and its definition. While ASH rates have traditionally tended to focus on primary care provision (particularly in general practice), it is important to remain aware that many other aspects of the health care system – hospital supply and configuration, emergency care department management, community care provision etc – can have an effect on ASH.

For these reasons, both the Ministry of Health ASH and a modified version of ASH are presented. ‘Modified ASH’ allows people to further interrogate this dataset by removing filters and exclusions normally applied to ASH data. The removal of filters and exclusions allows users to see all ASH events, including emergency department admissions and non-casemix events. Weighting was not applied as presenting all events was thought most important - the relative preventability of any ASH event is worth consideration. The removal of filters and exclusions increases the numbers reported and alters the ranking of DHBs markedly for overall ASH rates and angina. To view methodology click here.

It is important to note the deliberate use of the word ‘sensitive’ in the title of ASH – not all these admissions would be preventable even in a perfect health system. Moreover, many ‘unplanned’ admissions are planned in the acute sense by the primary care clinician, and are not ‘avoidable’ given current health service resources, and the psycho-social circumstances of patients and their families.

It is also important to think clearly about the focus of any apparently preventable admission. Angina is an example where ‘preventability’ could refer to acute primary care interventions, such as a decision to seek an opinion or not, or much longer-term public health measures, such as smoking cessation legislation.

For all of the conditions in the Atlas, many of the admissions are warranted using present health system conditions. Hence, zero is not the sought target here.

Typically ASH reports do not include the 75 – 84 year age group. This group was included in this analysis because it was considered important to encourage debate as to how older people are cared for and particularly in relation to rest home care and hospital length of stay.

It has been suggested that to interpret ASH analyses, the preventability of classes of admission through access to specific primary care interventions should be considered, rather than individual failures to prevent admissions of an individual. In reporting rates for older adults, two factors may impact on admissions:

  • individual frailty, social support and home situation
  • rest home bed availability.

Key messages:

ASH admissions represent 18–28% all acute/arranged admissions in people aged 75–84. Admission rates were significantly higher in older adults and in Māori and Pacific peoples at around 1 in 5, and 1 in 3, respectively.

There was a two-fold variation in admission rates for angina/chest pain and congestive heart failure, with some DHBs having consistently high rates.

Admission rates for bacterial/non-viral pneumonia were less variable with less than a two-fold variation.

The number of older people having two or more admissions a year for any ASH condition was analysed. Readmissions ranged from 39 percent of all ASH admissions up to 56 percent.

The double map includes possible drivers for admission rates, such as deprivation and the availability of aged residential care. It also allows the user to compare the effect of modified ASH and to investigate whether admission rates for people aged 15–74 due to angina/chest pain or congestive heart failure relate to admissions for the same conditions in older people.

In reporting these rates, the intention is not to prevent all admissions but to question:

  • How might high rates of admissions be affected by more intensive support for primary care management?
  • How do health literacy, service provision, primary care access and prevention programmes impact on admission rates?

Recommended reading

Baird J, Smith P. 2011. A short report on integrated care initiatives in selected New Zealand Health Networks. URL:

Basu A, Brinson D. 2008. The effectiveness of interventions for reducing ambulatory sensitive hospitalisations: a systematic review (Vol 1). University of Canterbury: Health Services Assessment Collaboration.

Kruzikas D et al. 2004. Preventable Hospitalizations: A Window into Primary and Preventive Care. HCUP Fact Book 5, 0056.

Purdy, S. (2010). Avoiding hospital admissions. What does the research evidence say? The King's Fund.




  1.  Jackson G, Tobias M. 2001. Potentially avoidable hospitalisations in New Zealand, 1989-98. Aust NZ J Public Health 25(3): 212–21.
  2.  Kruzikas D et al. 2004. Preventable Hospitalizations: A Window into Primary and Preventive Care. HCUP Fact Book 5, 0056.

Last updated 03/10/2016