Monitors do not always provide an accurate reading, particularly when there is respiratory artifact on the waveform. In order to get the most accurate reading, CVP should be read during the end-exhalation phase of breathing, when intrathoracic pressure is "zero". That is, there is not negative pressure for inhalation or positive pressure during exhalation (or vice versa on the ventilated patient).
The CVP waveform consists of three waves (A, C, V). The 2 most accepted ways to obtain a CVP reading are to take the mean of the A wave, or find the z-point. Because determining which wave is the A wave can be difficult, especially for newer nurses, and because of abnormal rhythms (a-fib, AV dissociation) and physiologic abnormalities , I think using the z-point method is easiest.
The z-point is found at the end of the QRS, just before the C wave (which represents tricuspid valve closure). It is a good indicator of right ventricular end diastolic pressure. To do this, make sure your transducer is at the phlebostatic axis and zeroed to atmospheric pressure. The head of the bed can be up to 30 degrees. Print a strip of your ECG and CVP waveforms, determine exhalation on the respiratory cycle, taking into account whether your patient is spontaneously breathing or on a vent (remember that exhalation is almost always longer than inhalation). Draw a line from the end of the QRS down to the CVP waveform and take the reading (be sure to check the scale!) - best to average a few. When this is done, you shouldn't have a negative number which is usually caused by negative intrathoracic pressure.
All this said, many clinicians simply trend CVP measurement and treat it (or end treatment) when it has changed significantly. However, if there is significant respiratory artifact, you could be over- or under- treating your patient based on a "false" reading.
Hope this helps and doesn't confuse things!