Posted by GEF on

Extracting meaning from numbers

One function of SortED tablet has not really been put to the test during our simulated clinical-use trials in the EDs at Charing Cross and St Mary’s, or in the UCC at Charing Cross Hospital.

 

We’ve focused on the main functions, but there is an interesting part of the toolkit we have yet to put to the test with real patients.

 

Venous and arterial blood gas analysers are increasingly used for point-of-care testing (POCT). Other POCT devices such as iStat Chem 8+ provide a somewhat different range of tests.

 

These anaylsers print out paper strips, usually with about 20 numeric values (for measurements (pH, pCO2, pO2, Lactate etc.) and derived figures like base excess.  These numbers are difficult to interpret even for some clinicians. Ultimately will use Bluetooth, or similar techniques, to port the results to SortED, but in the meantime we use  similar one-click data entry panels to those for basic observations. They are rapid to fill, but so that we can deliver interpretations, we have made a suite of interpretation algorithms. These not only interpret the findings into simple clinical terms, but also, where abnormalities are significant, the findings contribute to acuity scoring.

 

The messages to the user either interpret, advise a further computation, or make a cautionary comment about a reading. Most types of acid-base disturbance can be interpreted as either respiratory, metabolic or combined and as compensated or pure type. Other metabolic abnormalities or abnormal results also trigger messages (and add them to the patients’ SortED report.

 

  • The pH result ‘Moderate acidaemia,’ might be accompanied by ‘metabolic acidosis (pure)’  or ‘respiratory acidosis (compensated).’
  • A pair of blood urea nitrogen and creatinine entries might elicit the message ‘Check BUN/Creatinine ratio. If < 10 suspect renal failure’
  • Recording a very high potassium would elicit a caution message ‘Consider artefact due to haemolysis of the sample’
  • Raised carboxy haemoglobin entries (from carbon monoxide poisoning) trigger the warning  ‘COHb levels do not correlate well with severity of symptoms/signs.

 

We now need a collaborative study with an ED with high volume use of its VBG/ABG to give this system a thorough workout and an independent expert to evaluate how well our suite of algorithms gets on.

 

Gillie Francis – Aug 2018