In the 11th part of our series, Richard Marshall, Hugh Montgomery, David Howell, and Neil Goldsack explain how to understand and interpret common bedside monitoring

The allure of the gadget is always there for a doctor. It's just so tempting to rattle off a test or connect the patient to a machine that goes ping! We feel safer with our own clinical judgment if we can "put a number on it." But just how useful are these machines and tests in acutely unwell patients? We hope that by now we have hammered home the importance of applying basic clinical skills in acutely unwell patients, in particular the examination. Despite this dictum, bedside monitoring is often helpful, and further tests are usually required to assess the extent of the problem more accurately and to guide treatment.

But are they to be trusted? What information do you need to have to interpret them successfully? In this month's article, we will outline some of the more simple bedside monitoring used, their principles and interpretation. In particular, we will highlight the potential dangers of overreliance on monitoring and the pitfalls in interpretation. This is not intended to deter the use of monitoring but to encourage you to be circumspect and perhaps brush up a little physiology (yes, you knew it had to come in useful one day!) If you remember nothing else of what follows, remember this: if the value a monitor or test gives you doesn't agree with your clinical judgment, be suspicious.

The blood pressure

This sounds simple enough. Slip a cuff on and away you go. The blood pressure is used for many purposes including an assessment of the circulating volume and to determine cardiac output among others. Yet the blood pressure measures neither of these directly, although it can give you clues. The blood pressure can be over- or underestimated by arm cuff. The human ears are almost deaf to the sounds needed to measure blood pressure (25-50 Hz for the Korotkoff sounds), and many studies have shown that we seriously underestimate the blood pressure in clinical shock by the cuff method. If a blood pressure is low with a blood pressure cuff, it is likely that the actual pressure is even worse, so act on it. Don't wait. Prolonged hypotension is a killer. Of course even when we do measure the blood pressure, we often do not place the value obtained in its correct clinical context.

For example, a house officer was recently called to see a patient who had had an infusion of frusemide (furosemide) and dopamine, whose urine output was 200 ml/hour and blood pressure 90/50. In theory, the patient's blood pressure was low. In practice he was well. They were treating the test, not the patient. Remember, a "normal" blood pressure is that which allows for normal organ function (brain, kidney, and heart are organs of prime importance). The patient above was a boy aged 17 years. His blood pressure was indeed 90/50 on an arterial line measurement, but this was perfectly adequate for him, and he did not require diuretics or inotropes. The explanation for this is partly that the main driving pressure for tissue perfusion is the mean blood pressure not the systolic pressure. You can estimate the mean blood pressure at the bedside by remembering that diastole is one third of the cardiac cycle therefore for the above patient, his mean blood pressure would be: (90×0.66)+(50×0.33)=59.4+16.5=75.9 mmHg

As a very general rule of thumb, a mean blood pressure above 60 mmHg is usually adequate for tissue renal perfusion. Conversely, don't be fooled by a "normal" pressure. Our house officer trudged off to see an oliguric patient with a blood pressure of 120/70. Frusemide (furosemide) and dopamine were again in evidence. In fact the patient had been previously hypertensive (average blood pressure 170/90). Such patients often require supranormal blood pressure for adequate tissue perfusion. Fluids sorted this patient out.

The central venous pressure

The central venous pressure is (in the absence of clinically significant valvular disease or superior venacaval obstruction) an index of right atrial pressure and is also related to the pressure in the right ventricle at the end of diastole (but not always). A few simple points to remember:

• There is a considerable normal variation in the central venous pressure (up to 5 cm H2O). This is partly the result of changes during the respiratory cycle. Changes of less than 3 cm H2O should therefore be considered insignificant.
• The venous system is a capacitance system by nature. Veins are very compliant, which means that large changes in volume are absorbed by the system with little change in pressure. When the system has reached "capacitance," however (ie, it's full!), the central venous pressure can suddenly rise.
• Vessel compliance varies with age so a normal central venous pressure is only normal for that person. Often the response to a treatment - for example, giving fluid - is more informative than a single measurement. For example, a hypotensive patient was sound to have a central venous pressure of 16 cm H2O. This was misinterpreted as meaning that the volume of fluid was normal. In fact, 1 litre of fluid later, the central venous pressure had only risen by 1 cm H2O, suggesting there was still considerable "space" for more fluid. The patient was elderly, and his venous tone was high.
• Of course, fluid is not the answer in all circumstances and can be potentially harmful in patients with left ventricular failure. If, however, you suspect that a low circulating volume is the problem (in haemorrhagic shock, for example) do not accept a single central venous pressure measurement as normal if it does not agree with you clinical assessment.

Box 1 - Pulse oximetry Beware SaO2<95%. This is usually abnormal Only changes >3% should be considered clinically significant Don't take the oxygen off to see what happens to the SaO2 Oxygen treatment can mask a fall in PaO2 while maintaining SaO2 Beware carbonmonoxy-haemoglobinaemia

Pulse oximetry

Pulse oximetry almost certainly represents the most important step forward in patient monitoring in the past 20 years. The machines are non-invasive, comparatively inexpensive, portable, and should be available wherever there are likely to be acutely unwell patients. You can use fingers, toes, or ear lobes. The principle is straightforward. Light at two wavelengths passes through a blood vessel. Haemoglobin preferentially reflects the light at one wavelength (940 nm) oxygenated haemoglobin (HbO2) at the other (660 nm). The proportion gives the percentage of Hb that is fully saturated.
 

The haemoglobin-oxygen dissociation curve (remember that 1 kPA=7.6 mmHg).

Pitfalls

• Oximeters are accurate to within 3%. In the absence of chronic lung disease, the oxygen saturation (SaO2) should be above 95%. Below this is abnormal and a cause should be sought.
• Patients with chronic lung or cardiac disease may have saturations below 95%, but be very careful. Assume these values are abnormal until convinced otherwise.
• Take care if the patient is peripherally cold and "shut down" as this will reduce measured oxygen saturation but may not reflect central SaO2. Often, however, the peripheral cooling is the result of cardiorespiratory failure and the SaO2 is genuine.
• Beware blue or red nail polish. It affects the reading.
• Finally, remember that carbon monoxide haemoglobin will reflect as much red light as HbO2. Oximetry is therefore unreliable if there is significant carbon monoxyhaemoglobinaemia or methaemoglobinaemia.

Box 2 - Blood gas analysis Do not take the oxygen off to obtain a blood gas analysis Hypoxia is much more likely to be a problem in COPD that a raised CO2 Watch the base excess (+)/deficit ( ) Thus, a PaO2 of 16 KPa on 60% is very abnormal

Interpretation

Remember your oxygen haemoglobin dissociation curve. At the upper plateau of the curve, oxygen saturation is maintained in the face of a falling partial pressure of oxygen (PaO2). At a certain point, however, the saturation will fall precipitously. Consider the following common scenario. A patient with asthma was admitted to the ward. An oximeter was in place, and the SaO2 remained at 95% throughout the night until "suddenly," at 6 00 am, the SaO2 plummeted to 77%. What had happened so suddenly? Nothing, is the answer. The patient had been given 60% oxygen all night, and this had masked the gradual fall in PaO2. The lesson is to be wary that oxygen treatment will maintain SaO2 despite large changes in underlying oxygenation. The solution is to be sensitive to even small changes in SaO2 (>3%) and consider formal blood gas analysis to measure the PaO2. Do NOT take the oxygen off "to see what happens to the saturations." This is a dangerous practice. Do a blood gas analysis.

Box 3 - Central venous pressure The central venous pressure can normally vary by 3-5 cm H2O Only changes greater than 3 cm H2O should be considered clinically significant A "normal" central venous pressure is only normal for that person The response of the central venous pressure to therapy tells you much more than a single measurement

Blood gas analysis

There is not enough space to adequately cover the whole of blood gas analysis. There are a few pitfalls and misconceptions. When a patient is being treated with oxygen, do not stop giving oxygen before performing blood gas analysis. This is dangerous. The predicted PaO2 can be estimated from the alveolar gas equation, but you're unlikely to remember this. A useful rule of thumb is that the PaO2 should be the inspired oxygen percentage minus 10-15 KPa. Thus: On 60% oxygen the PaO2 should be 45-50 KPa. A PaO2 of 16 KPa on 60% is therefore very abnormal.

In patients with chronic obstructive airways disease, oxygen treatment is often withheld because of a concern over CO2 retention. This is the result of a loss of the normal respiratory drive to increased CO2. This phenomenon is in fact uncommon, and many patients with severe hypoxia are deprived of essential oxygen treatment on this basis. Of course you should be aware of the potential for CO2 retention and recheck the blood gases after starting oxygen treatment. But give the patient oxygen.

Hypoxia is much more likely to cause problems in these patients than CO2 retention. The base excess is often ignored on a blood gas analysis. This is a calculated value indicating the excess or, more often, deficit (a negative number) of basic ions. Its normal range is +2 to 2. A worsening base deficit (becoming more negative is a sensitive indicator of acidosis. It should prompt you into action. In particular, consider any cause of an acidosis. Think hypotension (with poor tissue perfusion), particularly fluid depletion, think renal failure, and think sepsis. A base deficit >5 suggests a serious acidosis is developing. The pH may be normal because of compensation, so beware and take note of this number.

Urine output

As a rule of thumb, a normal person should pass urine equivalent to 0.5-1 ml/kg/hr. A person weighing 70 kg will therefore be expected to pass 35-70 ml/hr. Less than 30 ml an hour is clearly abnormal, and you should look for a reason. Make sure the urinary catheter is not blocked. Think fluid (blood pressure, measurement of central venous pressure if required). Don not what to see what will happen in someone with oliguria, you need to do something immediately. Watch his/her base deficit. If s/he is worsening call expert help.

Conclusions

Normal is only normal for an individual patient.

The value you obtain should form part of a "picture" based on clinical examination, history, and the results of all investigations.

A repeat test after treatment often tells you much more than the initial test. If the value a monitor or test gives you does not agree with your clinical judgement, be suspicious!

Understand how a test or monitoring value is obtained and how it relates to the underlying physiology.