The links below will take you to QSM information for the July – September 2014 quarter.

Central line associated bacteraemia

The very high level of compliance with the insertion bundle continues. Nationally, 95 percent of insertions were compliant with the bundle, and all intensive care units bar one have more than 90 percent of lines inserted in compliance with the bundle.

Percentage of intensive care unit central line insertions fully compliant.

As the following run chart shows, CLAB infections continue to be rare, at well under 1 per 1000 line days.

Central line associated bacteraemia infections per 1000 line days

The consequence of this continued low CLAB rate is that in the two and half years since March 2012, 234 CLAB infections have been avoided at a saving of over $4.7 million.

Central line associated bacteraemia and associated costs since 2012.


Nationally, risk assessments of older patients* remain broadly consistent (89 percent in quarter three compared with 89 percent in quarter two). Three-quarters of district health boards (DHBs) assessed 85 percent of older patients or more in quarter two of 2014, indicating that nearly all DHBs are risk assessing nearly all older patients.

*Patients aged 75+ (55+ for Māori and Pacific peoples)

Percentage of older patients assessed for the risk of falling.

(Note: for this update, baselines and the most recent two quarters are shown for all tables.)

Patients at risk of falling are significantly more likely to receive an individualised care plan (now 90 percent, compared with 80 percent at baseline). However, there remains a tail of DHBs with lower care planning rates.

Individualised care plan that addressed falls risks.

When assessments and care plans are plotted against each other, there continues to be an increase in the number of DHBs in the ideal box of high assessment and high care planning (the top right corner) of the following graph.

Falls assessment compared with care planning.

In terms of harm reduction, when 2013–14 is compared with the baseline calendar year (2012), there has been a small, but as yet statistically insignificant reduction in fractured neck of femur cases. The major change is the sharp reduction in the additional occupied bed-days associated with these patients. This is the effect of there being a few very long stay cases in early 2012 which were not repeated. Based on a fixed cost for a fractured neck of femur, there was only a modest cost saving in 2013 of just over $100,000, but based on a length of stay cost estimate, this represents a reduction of nearly $2 million.

Harm and cost associated with in-hospital falls.

However, these figures disguise a disproportionate increase in admissions for older people most at risk of falling since the baseline period. Based on this increase, there have been in fact 11 fewer falls with fractured neck femur in hospital in the 16 months since the Open for better care campaign started (May 2013) than would have been expected had the rate of falls remained the same.

Looking at this month by month to identify a trend, falls with a fractured neck of femur remain fairly consistent at 6–12 incidents a month for most months of the last three years, with a median value of 8. In the last year, eight months have had a lower than average number of falls, but there has not yet been a clear run of six months below the median, indicative of a significant sustained shift.

In-hospital falls with fractured neck of femur by month.

Hand hygiene

The national compliance with the five moments for hand hygiene is now 75 percent, a level achieved by 15 DHBs. All DHBs bar one are now undertaking enough observed moments.

Percentage of opportunities for hand hygiene taken July–October 2014.

DHBs that submitted fewer than 50 percent of the required moments.

*The hand hygiene audit requires a minimum number of moments to be observed in order for robust data to be gathered. In the November 2013 to March 2014 audit period, five DHBs did not observe the required number of moments and these are shown separately. Canterbury and MidCentral both submitted the required number of moments prior to quarter one, 2014.

After a short-lived increase of Staphylococcus aureus rates between April and June 2014, the national rate is now back to its long-term average of around 0.12 per 1000 bed-days.

Recorded S aureus bacteraemia per 1000 bed days.

Medication safety – new measure

This quarter we introduce a quality and safety marker in a new topic area, medication safety. In general we construct quality and safety markers from process measures (ie, following a specific action designed to improve quality) and outcome (ie, the results of this). However, we are adopting a slightly different approach for medication safety.

Our focus is on the task of medication reconciliation – a process by which health care professionals ensure all medicines a patient is taking are known and reviewed to ensure they are appropriate and safe. Doing this reduces the risk of medicines with potentially dangerous interactions being prescribed.

A key first step to allow this to be done routinely and in a more straightforward manner is the introduction of an electronic system, known as electronic medication reconciliation (eMR). Currently a minority of DHBs have access to the system. There is a national programme to roll it out around the country.

Our first measure, therefore, is structural and looks at national implementation of eMR. The measure shows that four DHBs have implemented the system and a further three plan to implement it in the first quarter of 2015.

In future quarters this measure will be supplemented with process measures looking at the proportion of older patients (those, typically, likely to benefit most from eMR) for whom reconciliation was undertaken.

STRUCTURE: Implementation of eMR

Implementation of electronic medication reconciliation.

Perioperative harm

The increase in the recorded use of the surgical safety checklist continues, with the national rate now at 94 percent – an increase of nearly 25 percent from the baseline. Nearly all DHBs achieved the threshold level, or were not significantly lower than it. However, the paper-based method of data collection for this measure may be contributing to an overly compliance-focused use of the checklist, rather than enhancing teamwork and communication among surgical teams, as it intends to do. Some DHBs are trialling a paperless approach to the checklist (using a poster on the wall to prompt discussion). The Health Quality & Safety Commission is currently considering options to adapt the data collection for this measure, and will consult with DHBs on potential new approaches next year.

Percentage of operations where all three parts of surgical safety checklist used

There is now enough data to estimate how many cases of postoperative deep vein thrombosis (DVT)/pulmonary embolism (PE) and sepsis were avoided in the last two years in comparison with the baseline year of 2012. Our calculations suggest there have been 51 fewer postoperative cases of DVT/PE in this period but 113 more cases of postoperative sepsis than would have been expected using 2012 rates. However, the additional hospital stay associated with each DVT/PE and sepsis case has fallen since 2012, resulting in 1400 fewer bed-days associated with these cases. Across the two complications this is a cost of $1 million saved. We are working to refine this calculation by reflecting the relative risk of each case.

Harm and costs from DVT/PE and sepsis following operation from Jan 2013.

Surgical site infection

As the Commission uses 90-day outcome measures for surgical site infection, these data run one quarter behind other measures. Information relates to January–March 2014 (quarter one).

Process measure 1 – antibiotic administered in the right time
For primary procedures, an antibiotic should be administered in the hour before the first incision (‘knife to skin’). This should be happening in all cases, so the Commission has set the threshold at 100 percent. The number of DHBs achieving 100 percent in the last quarter rose to four. Thirteen DHBs have rates of 95 percent or higher, compared with just five DHBs in the baseline quarter. As before, non-achievement of the 100 percent target reflects failure to record timing of antibiotic administration rather than recorded incorrect timing. In fact, nationally, 98 percent of operations where an administration time was recorded had the correct time window.

Antibiotic given 0–60 minutes before knife to skin.

Process measure 2: Right antibiotic in the right dose – cefazolin 2g or more
Cefazolin 2g is recommended for routine antibiotic prophylaxis for hip and knee replacements unless the patient has a beta-lactam allergy and requires a non-beta-lactam antimicrobial agent, or is colonised with multi-resistant S aureus (MRSA), in which case they should receive both cefazolin and vancomycin. To allow for these relatively rare instances, the threshold is set at 95 percent. The use of 2g or more of cefazolin continues to increase, with 10 DHBs reaching the threshold level compared with only six in the previous quarter and three at baseline. In nearly all DHBs where this level was not met, the issue was of a lower dose (1g) of cefazolin being given. The exception is MidCentral, which uses other antibiotics.

2 grams or more cefazolin given.

Process measure 3: Appropriate skin antisepsis in surgery using alcohol/chlorhexidine or alcohol/povidone iodine
Skin preparation involving either chlorhexidine or povidone iodine in alcohol should occur on all occasions, so the threshold is set at 100 percent, a level now achieved by 13 DHBs compared with eight at baseline. Appropriate skin preparation is clearly normal practice across DHBs, as the national compliance rate of 97 percent attests, an increase from 91 percent at baseline.

Appropriate skin preparation.

We measure infection rates within 90 days of the operation, meaning that we report outcomes a quarter behind other measures. After a reduction in infection rates between February and April, the most recent months have seen infection rates return to the long-term trend of between 1.2 and 1.3 infections per 100 operations.

Surgical site infections per 100 hip and knee operations.

Last updated 27/10/2015