5. Scoring Systems

early identification of clinical deterioration and to prevent delayed intervention. It is a simple ward-based scoring system that is determined from the standard observation chart. A score of 0, 1, 2 or 3 is assigned to the six routine physiological variables of respiratory rate, heart rate, systolic pressure, temperature, urine output and neurological status score is therefore 0 and the highest possible is 18. A score of 4 or more usually triggers referral initially to an appropriate medical professional, usually the ward doctor. There are many local variations.

MEWS. PEWS is the Paediatric Early Warning System used in patients below the age of 16 and in whom the reference ranges are quite different, particularly in much younger children charts for infants, pre-school (1–4 years), school age (5–12 years) and teenage (12–15 years). Typically, PEWS looks at ‘Respiratory’ (rate, recession, accessory muscles, tracheal tug), ‘Cardiovascular’ (colour, capillary refill, tachycardia of >30 above normal rate) and ‘Behaviour’ (alert, sleeping, irritable, lethargic

assessment of conscious state based on eye opening (1–4), verbal response (1–5) and motor response (1–6), giving a minimum score of 3, either deep coma or death, and a maximum of 15, which is full consciousness. It is included as part of several other scoring system the AVPU score which determines whether a patient is Awake, responding to Voice, responding to Pain, or Unresponsive

Wilson score (prediction of difficult intubation). This is a simple tool for airway assessment with a point assigned to each of five variables: weight, craniocervical mobility, mouth opening, retrognathia and prominent overbite. It has minimal interobserver variation and so is more reliable in that respect than the Mallampati test, but although the sensitivity of the Wilson scoring system is high, its specificity is low.

Apfel score (risk of postoperative nausea and vomiting). The Apfel score estimates the likelihood of PONV according to the number of predisposing factors. These are previous history, non-smoking, female gender and administration of opioids. The incidence of PONV in relation to the number of factors is quoted as 0–10%, 1–20%, 2–40%, 3–60% and 4–80%. Clinical experience suggests that this would seem improbably high, but validations of the various models suggest that they can be useful.

estimates mortality based on laboratory values on admission with some repeated at 48 hours. It gives one point to each of the following 11 values: white blood cell count>16,000; age >55 years; glucose >10 mmol/L on admission; aspartate aminotransferase (AST) >250 i.u. l–1 ; lactate dehydrogenase (LDH) >350 i.u. l–1; at 48 hours: haematocrit fall of >10%; increase in urea concentration by >1.4 mmol l–1; serum Ca2+ <2.0 mmol l–1; base deficit >4; fluid requirements of >6 l within 48 hours; PaO2 <8 kPa. A total of 5–6 points predicts a mortality of 40%; at 7–11 points this rises to 80%. (Ranson scoring has been used for several decades, but the Computed Tomography Severity Index appears to be a better predictor).

Model for End-Stage Liver Disease: MELD scoring. This is another organ-specific scoring system that can be used to assist decision-making around liver transplantation. The formula is complex, although it only uses three indices: bilirubin concentration, prothrombin time and creatinine concentration (3.78  [bilirubin] + 11.2  INR + 9.57  [creatinine] + 6.43), and it yields a round number which allows a prediction of 3-month mortality in hospitalized patients. If the score is 40 or more, for example, predicted mortality is greater than 70%.

Model for End-Stage Liver Disease: MELD scoring. This is another organ-specific scoring system that can be used to assist decision-making around liver transplantation. The formula is complex, although it only uses three indices: bilirubin concentration, prothrombin time and creatinine concentration (3.78  [bilirubin] + 11.2  INR + 9.57  [creatinine] + 6.43), and it yields a round number which allows a prediction of 3-month mortality in hospitalized patients. If the score is 40 or more, for example, predicted mortality is greater than 70%.

POSSUM and P-POSSUM. This is the Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity. P-POSSUM The scores are calculated using an equation which combines weighted variables both of physiological and operative data. These are too detailed to be discussed in detail in the oral, but the 12 physiological variables consist of age, cardiac signs, respiratory history, pulse rate and systolic blood pressure, Glasgow Coma Score, haemoglobin concentration, white blood cell count, urea, sodium, potassium and the ECG The six operative variables are severity, multiple procedures, malignancy, peritoneal soiling, total blood loss and the CEPOD mode of surgery (elective, urgent, emergency).

POSSUM and P-POSSUM. This is the Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity. P-POSSUM The scores are calculated using an equation which combines weighted variables both of physiological and operative data. These are too detailed to be discussed in detail in the oral, but the 12 physiological variables consist of age, cardiac signs, respiratory history, pulse rate and systolic blood pressure, Glasgow Coma Score, haemoglobin concentration, white blood cell count, urea, sodium, potassium and the ECG The six operative variables are severity, multiple procedures, malignancy, peritoneal soiling, total blood loss and the CEPOD mode of surgery (elective, urgent, emergency). combined into a formula which subsequently was judged to overestimate the risk of mortality, particularly in lower-risk patient

There are multiple critical care scoring systems of which APACHE is the most widely used. However, their function can be misunderstood. They are not intended to allow a prediction of morbidity and mortality in an individual patient, tempting though it may be to use them in that way.

This is the Acute Physiology And Chronic Health Evaluation score which has had further iterations since its introduction in 1981 number of variables from 34 to 12 and altered some of the weightings, for example, giving higher weightings for acute kidney injury. during the first 24 hours of admission to critical care. Each variable is scored between 0 and 4; a total of 25 is predictive of 50% mortality, which rises to 80% with a score of 35 or greater These are temperature, mean arterial pressure, heart rate, respiratory rate, A-a DO2 (FiO2 > 0.5) or PaO2 (FiO2 < 0.5), arterial pH, serum Na+, serum K+, serum creatinine, haematocrit, white cell count and Glasgow Coma Score

SOFA. Sequential Organ Failure Assessment. SOFA takes the same six variables and scores the worst values on each day, usually for 72 hours. The main difference between MODS and SOFA is that the latter includes therapeutic measures, whereas MOD scoring is independent of treatment. The two are similar in respect of reliability. (Respiratory: PaO2/FiO2; Cardiovascular: HR  CVP/MAP; Renal: [creatinine]; Hepatic: [bilirubin]; Haematological: [platelet]; Neurological: Glasgow Coma Score).

ICNARC. This is the Intensive Care National Audit and Research Centre which in effect was established following a national study of APACHE II scoring in UK critical care units. It describes itself as a national centre for comparative audit and research in intensive care, in the context of which the various scoring systems and their evaluation have obvious relevance. ICNARC also looks at data such as the standardized mortality ratio. This is the ratio of the observed mortality rate to the expected mortality rate and takes into account the severity of illness (using scoring systems as described previously) and the differences in case mix