Kvalitets scoring

Vi vet at det å gi god HLR kan bidra til høyere overlevelsesrate. Ideen med å score en HLR prestasjon er å gi en indikasjon på hvor god behandling pasienten får. 

Tidligere har HLR tilbakemeldingen vært godkjent/ikke godkjent. I løpet av de siste ti årene har vi i Laerdal sammen med partnere laget en algoritme som gir en mer detaljert og granulær poengsum. 

Eksempel: Møt Julie. Hun jobber med å utøve HLR og mislyktes når hun tok HLR testen. 

Julie er akuttsykepleier på et sykehus i England. Hun er i fysisk god form og i stand til å gi solid dybde på mer enn 5 cm på hver kompresjon. Hun har god oppslipp og ingen avbrudd. Hun er en av de beste i å gi HLR på sykehuset, alikevel var den binære tilbakemeldingen at hun hadde mislyktes. Hvorfor? 

Hastigheten på kompresjonene hennes var 122 kompresjoner per minutt. Klinisk suverent, men litt uten for retningslinjen som er 100-120 kompresjoner i minuttet. Er det rettferdig å fortelle henne at hun hadde mislykkes? Nei. Julie, kunne bidratt til å redde liv. 

Vi så disse resultatene hele tiden når vi brukte den gamle, binære scoringen. Dette er en av grunnene til at vi laget en ny algoritme for HLR scoring. 

Mens det er vanlig å kun bruke binære tall, bruker vi i Laerdal en ikke-binær tilnærming for å presentere en mer realistisk scoring. Den enkle grunnen er at selv om både 49 mm og 25 mm kompresjonsdybde er utenfor retningslinjen , er 49 mm utvilsomt en mer ønskelig dybde.

• Binær scoring: I det innledende eksempelet hadde elev 1, 94% av kompresjonene innen riktig dybde 50-60 mm. De resterende kompresjonene lå utenfor retningslinjene. Det skilles ikke mellom for eksempel 25 mm og 49 mm dybde, begge beregnes som feil. 
• Ikke-binær tall:  Jo større avvik, desto lavere score. Men andre ord, 49 mm kompresjonssybde gir en bedre poengsum enn 25 mm dybde. 

Figur 1. Ikke-binær scoring
Elv 1 og 2 har like mange kompresjoner innenfor og utenfor retningslinjene. Men elev 2 får høyere poengsum siden kompresjonene som er utenfor retningslinjen er nærmere retningslinjen. Det samme prinsippet brukes på alle andre HLR-ferdigheter. (figuren er kun for illustrasjon)

The algorithm which makes up the QCPR score is made by Laerdal Medical in close collaboration with members of the AHA ECC Subcommittees and co-authors of the 2013 AHA Consensus Statement on CPR Quality. Based on input from these specialist members, we created mathematical models for each sub-skill in CPR performance, like compression depth, leaning, ventilation volume, etc.

Each compression and ventilation are tracked and scored individually and summed up in the overall score. Even though numbers like average rate can be interesting, averages are not used to calculate the overall score. Similarly, the binary numbers do not affect the score calculation directly.

Subtractive scoring model
The scoring algorithm use a subtractive scoring model where we start off on a perfect score and, if the learners do something wrong, the score is reduced. If you are within the guidelines you will always get a 100% score. If you deviate, the score is reduced.

Plotted along an S-curve, we can see that small deviations give small reductions in the overall score, while large deviations result in large reductions in CPR scores:

Fig.2 : Compression rate on the x-axis, and score on the y-axis.
We see that the CPR score deteriorates quickly on both sides of the guideline thresholds. (graph not to scale, for illustrative purposes only)

A popular feature of our QCPR manikins is the QCPR race where up to six learners can participate in a race to see which one performs the best CPR. 

Sometimes we see that while one participant achieves the best score on binary sub-metrics (like adequate depth percentage), another participant will be crowned the winner of the race. The binary sub-metrics (like percentage of adequate depth) will not always help explain the CPR score.

Example 1 - high score with low binary metrics: 

Two learner perform 100 compressions. Learner 2 gets a better overall score than learner 1, even though learner 1 had better compression depth if you only look at the binary metrics.

The reason is that learner 2 was closer to the guidelines during the 10 compressions that were not fully "adequate", and learner 1 was very far away from the guidelines during the 8 compressions that were not "adequate".    

Example 2 - Ventilations over the limit: 

In another example, both learners provide too much air in ventilation and miss the ventilation sub-metric.

Let's say learner 2 gave 610 ml of air (which is clinically OK) and learner 1 provided 2000ml of air, which indeed can compromise the clinical outcome. In binary terms, these results are the same, simply "not adequate". However, in clinical terms, and therefore also in scoring terms, they are very different. Learner 2 was much closer to the guidelines and receives a higher score. 

Sometimes - especially for 'Compressions Only' sessions - the total score is unexpectedly low, even though compression rate, depth, and release are all perfect (see image below).

The score reduction for many of those sessions comes from the chest compression fraction. If your compressions are interrupted, there will be a reduction in the score - longer interruptions give larger score reductions.

Also worth noting, if you are running a ”Compression only” sessions, the expected chest compression fraction is 100%, as you have no good reasons to stop the compressions. So, if there are interruptions, the score is reduced rather strictly.

Illustration, Compression only session: Rate, depth, and release are all perfect. However, the total score is unexpectedly low. The culprit is the chest compression fraction parameter. 

The number of parameters scored depends on which manikin or simulator you use, and which software or application you use. 

10 parameters used to calculate QCPR score:

• Compression depth
• Compression rate
• Incomplete release
• Hand position 
• Compression per cycle
• Chest compression fraction (also known as flow fraction)
• Ventilation volume
• Ventilation rate
• Number of pre-ventilations
• Inspiration time in pre-ventilations