AIM To investigate the range of activities carried out by pharmacists responsible for services to intensive care units (ICUs), and to look at patterns of drug expenditure and drug use.
DESIGN Postal questionnaire.
SUBJECTS AND SETTING Designated ICU pharmacists throughout the United Kingdom.
METHOD Responses were analysed to identify common roles of pharmacists in ICUs and the drugs most commonly used for analgesia, sedation and paralysis. Drug expenditure and patient activity were also compared and the possible links between these factors and pharmacist activity assessed.
RESULTS Pharmacists were commonly involved in clinical, budgetary and educational activities. There was a high level of pharmacy input in units with a high total therapeutic intervention scoring system (TISS) score, but not necessarily with a high drug expenditure. Morphine, midazolam and atracurium were the drugs most commonly used for analgesia, sedation and paralysis, respectively.
CONCLUSION A baseline understanding of pharmacy practice on ICUs in the UK was obtained. There is a need for more specific studies related to pharmacists' ICU activities and ICU drug expenditure.
Intensive care medicine has developed into a specialty in its own right, to the extent that there now exist accredited specialist training schemes for medical and nursing staff. The benefits of a multidisciplinary approach to patient care have been recognised in many areas,1 and there is no reason why this should not also be the case in the intensive care unit (ICU). For this to work, the professions involved must demonstrate the particular skills they can bring to the unit, and the practitioners must show that they can work effectively for the patients' benefit in this specialist environment.
Specialist pharmacists have developed in recent years to act as practitioners who are more focused in their areas of practice than "generalists". These specialists have a higher level of knowledge of the particular therapeutic interventions and disease processes encountered in their specialty and the experience to identify the possible interactions of clinical care with drug therapy. They can also liaise with the pharmacy department and the clinical department to ensure the most appropriate pharmaceutical services are provided. The benefits of specialist renal2 and psychiatric3 pharmacists have been documented.
The idea of a specialist intensive care pharmacist was first established in the pharmacy literature in 1982.4 The increasing use of high-tech drugs and sophisticated drug delivery methods would seem to make the ICU a prime area for specialist pharmacy practitioners. The progress of the specialty in the United States was documented in two papers published 10 years later,5,6 by which time national standards had been set and accredited training schemes developed.7 Montazeri and Cook showed the benefits of the pharmacist in a multidisciplinary ICU set-up in Canada,8 and cost savings from having pharmacist input to a medical ICU in the US have also been demonstrated.9 Recently, Leape and colleagues10 showed a lower rate of adverse drug events associated with pharmacist involvement on prescribing rounds in a medical ICU in the US.
In the UK, the specialty appears to have developed as the activity of pharmacists in ICUs has increased. This is shown by the increase in membership and activities of the UK Clinical Pharmacy Association's critical care practice interest group. Published information on the role of the pharmacist in intensive care in the UK, or of the impact on patient care, is limited,11 though Cadman and Park12 have shown clearly the potential for pharmacists to make drug administration optimal for ICU patients. However, the level of practice across the country as a whole is unknown.
Without a baseline knowledge of what is happening now, it is difficult for the specialty to progress on a national basis. The UKCPA critical care practice interest group has produced a leaflet regarding minimum standards for ICU pharmacists. This is based on the experiences of some leading practitioners rather than being evidence-based.
Unfortunately the Intensive Care Society's "Standards for ICU"13 make little mention of the need for pharmaceutical input. In order for this situation to be changed, the benefits of the specialist pharmacist must be proven, and the level of practice publicised. It was, therefore, felt that a survey of current practice, and collection of relevant information on drug use in ICUs, would be an appropriate first step in establishing a baseline.
The aim of the survey was to identify the extent of practice of ICU pharmacists, and to demonstrate some benefits of pharmacist involvement in critical care patient management. The study had three objectives:
A postal survey form was sent to 287 hospitals which had an ICU registered with the Intensive Care Society. Initially, this was sent to the ICU pharmacist at the hospital, but the response was poor. The survey was then addressed to the chief pharmacist at each hospital who had not responded to the first mailing, requesting that the survey be passed to the most appropriate person for completion.
The first section of the form included questions asking about the grade, qualifications and experience of the ICU pharmacist (if there was one) and how much time was allocated to the role. It then moved on to ask for information regarding the activities of the pharmacist on the ICU. The activities listed were based on those defined as "necessary" or "ideal" by Dasta6 but amended to take into account the different circumstances in the UK.
The form also gathered background information on the unit, such as number of beds and types of patient treated, before requesting information on the annual drug expenditure and patient activity. The final section asked for the first and second choice agents used for sedation, analgesia and paralysis.
There were 287 units on the original mailing list. Replies were received from 171 units, representing a 60 per cent response rate. In seven cases there was no ICU at the hospital. Of those hospitals that did have an ICU, 154 (94 per cent) had a pharmacist allocated to it. The range of times allocated to this activity each week is shown in Table 1. The mean time allocated was 9.3 hours per week (range 2–37 hours), or 1.4 hours per established bed per week (range 0.3–4.5 hours).
|
|
Ninety-seven per cent of ICU pharmacists were managed by the pharmacy department and the remainder (five pharmacists) were managed jointly by the pharmacy and the ICU. Respondents' grades are shown in Table 2. Fifty-six per cent of the ICU pharmacists had been qualified for 10 years or more, 35 per cent for five to nine years and 9 per cent for less than five years.
A postgraduate qualification in pharmacy was held by 75 per cent, and the most common ones were an MSc (38 per cent) and a diploma (32 per cent) in clinical pharmacy/pharmacy practice. Other qualifications included MPharm, MPhil and PhD and 55 per cent of respondents were members of one of the two main critical care interest groups.
The activities of pharmacists in ICUs are shown in Table 3. Activities categorised as "other" included research (four centres), clinical trials management, risk management assessment and a formal pharmacokinetic consultation service.
|
|
|
|
||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||
There was no particular geographical region or type of hospital (teaching/district general) that failed to respond to the questionnaire. The responses would, therefore, seem to provide a good representation of the situation in those hospitals where pharmacy involvement in intensive care is more than simply supplying the drugs requested by the nursing staff. However, the fact that responses were received from only 60 per cent of hospitals with ICUs, and that 6 per cent of the respondents had no pharmacy input to the ICU may mean that up to 50 per cent of ICUs have little input from pharmacists.
The fact that the survey was sent to hospitals registered with the Intensive Care Society does mean that a specific population was selected. Some may feel that those ICUs that are affiliated to the ICS are more likely to be actively looking at taking on new ideas and practices. Similarly a response may have been more likely from a centre with an active ICU pharmacist, so that the 154 ICU pharmacists may actually be almost all of the practitioners, not just 60 per cent as indicated from the survey. Moreover, the respondents represented a high percentage of the active membership of both critical care interest groups, though there is obviously still potential to increase membership.
One hundred and fifty-four (94 per cent) ICUs had a pharmacist specifically allocated to the ICU. Although this figure appears high, there was a considerable variation in the actual time spent on the unit with only 21 pharmacists spending 50 per cent or more of their time on the ICU; four were full time. It is not clear whether individuals were fully integrated into the running of the unit or whether their activity amounted to little more than a prescription monitoring visit. A mean of 1.4 hours per bed per week (or 17 minutes per day) suggests that the latter may be the case, as this amount of time would be unlikely to allow detailed analysis of drug usage information, development of protocols, involvement in training programmes and full involvement in clinical decision making.
When looking at the relationship between the time allocated and the expenditure and other activity indicators, the correlation was highest for the total TISS points (r=0.83). This may be because there were few complete returns, and it may have been those with a high involvement in ICUs who were most likely to be able to obtain information on TISS points. This suggests that the units with high activity (as indicated by TISS points) are more likely to have a higher level of pharmacy input. However, this needs to be confirmed by a more detailed study.
Most of the pharmacists responsible for ICUs "practised regularly" on the unit. This was a very loose description, intended to cover basic clinical activities, prescription monitoring and ensuring satisfactory medicine supplies, but it does indicate that most were comfortable with their involvement. Over half of respondents attend ICU ward rounds. Again, further investigation would be required to establish how frequent attendance actually is, ie, daily or weekly, as well as establishing the quality of the input on the rounds.
Provision of a formal adverse drug reaction (ADR) reporting scheme was low. However, this probably reflects the wider situation in that ADR reporting is not common practice in many hospitals, and positive responses may have come from ICUs in hospitals where there was a hospital-wide scheme rather than it being a unit initiative. This is disappointing, considering the evidence that pharmacy input into such schemes can improve their results,14 and the increased possibilities of ADRs being picked up in critically ill patients on multiple drug therapy.
Provision of drug utilisation review (DUR) and expenditure information was common, and almost three quarters of pharmacists were involved in providing training for ICU staff. Many of the new drugs and technologies are likely to be expensive compared to existing ones, so there may be an increasingly important role for the pharmacist to ensure scarce ICU resources are being used appropriately. Overall, the high level of activity described here suggests that many of the time allowances given (mean of 9.3 hours per week) may have been underestimates, perhaps relating to the time actually spent on the ICU, rather than the time spent working for the ICU.
The annual drug expenditure varied considerably, as would be expected from the wide range of size and specialisation of responding units (Table 4). Expenditure can, of course, be affected by different costing methods and differing local practices, such as whether certain items are supplied by pharmacy. The cost per bed, cost per patient and cost per patient day all show a relatively wide variation (Table 5). The last of these is possibly the most useful for comparison purposes as it removes the variables of bed use and extended stay patients.
The acute physiology, age and chronic health evaluation (APACHE) system is a tool for measuring the effectiveness and efficiency of the intensive care unit,15 and it is the second version of the scoring system - APACHE II - which is most commonly used at present. Many units produce a mean APACHE II score as a statistical indicator of activity over a period of time. As the APACHE score is an indicator of severity of illness, it might be expected that there would be a relationship between this and drug expenditure.
However, the figures obtained in this survey showed a relatively low correlation. This may be because some of the most severely ill patients do not live long enough to use a lot of expensive drugs, or it may reflect deficiencies in the data collection. Either way, it should probably be investigated further, possibly including total hospital length of stay and ultimate patient outcome, to see if it might produce a useful indicator for the level of drug expenditure for ICUs.
The therapeutic intervention scoring system (TISS) has been developed alongside APACHE and it attempts to measure the intensity of treatment on a cumulative daily basis by scoring points for various interventions. Thus, an annual total of TISS points for a unit might be related to drug expenditure on the basis that more activity would increase drug costs, and it would not be affected by early deaths. The data provided in this study do indicate a good correlation (r=0.77) between the total annual TISS points for an ICU and its drug expenditure. There were, however, only a small number of replies with usable TISS data, so the significance of this correlation may be lower compared with that of other parameters.
A more specific study of this area would be useful to demonstrate that the TISS is a better way of judging drug expenditure than the relationship to the average APACHE II score. Consistency of scoring for both APACHE and TISS would also be an issue as not all units complete statistics for all patients.
Only a moderate correlation (r=0.57) between drug expenditure and pharmacist time allocated to ICU was found. No relationship was observed between time allocated and either drug cost per patient day or drug cost per TISS point. This is disappointing, since one of the aims of pharmacists is to promote the cost-effective use of drugs. Of course, low cost does not necessarily equate with good patient outcomes, and patient outcomes would be a much more important measure. However, a pharmacist could easily spend a considerable amount of time monitoring the use of inexpensive drugs that would still have a considerable impact on patient care.
The use of infusions for drug administration was almost universal, with bolus administration being used mainly for paralysing agents. This is different from practice in the US, where a recent study showed only 51 per cent of doses were given by this method.16 However, there is little evidence to justify the use of continuous infusions. Park and Gray17 suggest that it is only convenience that has brought about the high use of this method of administration and that judicious use of bolus doses may cause fewer problems and allow better patient care. In some centres, however, nurses would need to be trained to administer the doses.
Morphine was the most popular first-line analgesic, with alfentanil being the second most popular first-line drug. Fentanyl was the most popular choice for a second-line agent. Altogether, these three agents accounted for over 90 per cent of all drugs used as analgesics in the intensive care units studied.
Midazolam and propofol accounted for virtually all of the sedatives used, with midazolam being the first-choice sedative for 50 per cent of respondents. The choice between these two agents has been a frequent subject for debate since it was suggested that, despite its higher cost, propofol could be more cost-effective than midazolam.18 More recently, it has been suggested that the use of bolus doses of lorazepam or haloperidol could be more cost-effective than either propofol or midazolam,19 but neither of these agents were mentioned by more than a handful of respondents in this survey.
Atracurium was clearly the first-choice paralysing agent, with vecuronium the usual second choice. Pancuronium, rocuronium and cisatracurium were also used, but in fewer cases. Atracurium was the only drug where bolus administration was significant, although of course this may increase the risk of tachycardia and histamine release if not done with care.
Having this information on drug choices will make it easier to compare and audit practices and find the most cost-effective treatments, as well as raising awareness of drugs that are used so that suitable presentations of the drug can be obtained commercially.
The low level of formal monitoring of drug action was worrying, particularly for paralysing agents. This may simply be that the pharmacists are not familiar with what does actually happen at the bedside, although part of pharmaceutical care should be the monitoring of drug effects.
All three study objectives were met. A good response to the survey was obtained, and a baseline knowledge of the level of pharmacy practice on ICU was obtained. This can be used as a start for future comparisons, as well as evidence for those wishing to develop their own service or personal practice. Interesting results were obtained on drug expenditure and also the relationship of expenditure to activity. However, this needs further investigation to get usable figures for benchmarking or planning purposes. Valuable information on the drugs used was obtained, and this part of the study could be repeated with other groups of drugs, such as inotropes and antibiotics, to help identify best practice.
ACKNOWLEDGMENTS The survey was aided by an educational grant from Aurum Pharmaceuticals Ltd. The advice and encouragement of Dr Peter Curry (consultant in anaesthetics and intensive care, Queen Margaret hospital) is gratefully acknowledged.
Mr Timmins is principal pharmacist, clinical services, at Queen Margaret Hospital, Whitefield Road, Dunfermline, Fife KY12 0SU