1. Medicines Reconciliation: Re-audit
Topic 8b
Medicines Reconciliation: Re-audit
The Prescribing Observatory for Mental Health (POMH-UK) runs national audit-based quality improvement programmes open to all specialist mental health services in the UK. The aim is to help mental health services improve prescribing practice in discrete areas (‘Topics’).
This slide set contains the results of the baseline audit for Topic 8: medicines reconciliation.
POMH-UK is based at the Royal College of Psychiatrists Centre for Quality Improvement and does not receive any funding from the pharmaceutical industry.
Further information about POMH can be found on our website:

2. Background Medication errors can occur at the point a medicine is:
prescribed (usually by a doctor)
dispensed (usually by a pharmacist)
administered (usually by a nurse)
Medication errors are:
a recognised common cause of avoidable morbidity and mortality
responsible for approximately 20% of clinical negligence claims in hospitalised patients (Audit Commission, 2001)
The point of transfer between care settings (in particular hospital admission) is a known period of high risk for prescribing errors (NICE, NPSA, 2007)
Errors can happen at the point a medicine is prescribed, dispensed or administered.
At each stage in the process the root cause may be a simple lapse in concentration, a problem with decision making, or a knowledge deficit.
Common types of prescribing errors include clerical errors made when transcribing medicine cards, and knowledge based errors which usually occur when prescribing unfamiliar medicines or medicines that have a high potential to interact in a clinically significant way.
Common types of dispensing errors include selecting a pack containing the wrong drug or dose, labelling with the wrong instructions, supplying the wrong quantity, or failing to identify a potentially serious drug interaction.
With respect to administration, the most common errors are selecting the wrong drug or dose, and giving the medicine by the wrong route or to the wrong patient.
Prescribing errors that are made at the point of transfer of care, particularly admission to hospital, are usually due to problems with medicines reconciliation.

3. Background Why this Topic was selected?
A technical patient safety solution that aims to reduce the prevalence of medication errors at the point of admission has been issued jointly by NICE and the NPSA. The safety solution focuses on the importance of medicines reconciliation.
The technical patient safety solution made it statutory that all Trusts had an approved policy for medicines reconciliation in place by December 2008.

4. Medicines Reconciliation
NICE/NPSA define medicines reconciliation as:
Collecting information on medication history
Checking this list against the current prescription chart
Ensuring any discrepancies are accounted for and actioned
Documenting any changes, omissions or discrepancies
“The aim of medicines reconciliation on hospital admission is to ensure that medicines prescribed on admission correspond to those that the patient was taking before admission. Details to be recorded include the name of the medicine(s), dosage, frequency, and route of administration. Establishing these details may involve discussion with the patient and/or carers and the use of records from primary care. This does not include medicines review”
(NICE, NPSA, 2007)
Errors occur most commonly on transfer between care settings, and in particular at the time of hospital admission.
While at home, a patient may be taking a combination of medicines, some of which may have been prescribed by a GP, some by a hospital specialist and some of which may have been obtained without the need for a prescription.
In addition, few people adhere completely to prescribed medication regimens; partial adherence and non-adherence are common. It is therefore possible, and indeed likely that not all medicines that are prescribed are being taken in the way intended. At the point of admission to hospital, medicines that the patient has been taking at home can be missed and medicines that are prescribed but are not being taken can be re-instated. Both have the potential to cause harm.
For example, medicines such as insulin that are used to treat serious physical health problems can be missed, and drugs with a dose related potential to cause postural hypotension, such as clozapine, can be restarted at a dose to which tolerance has been lost.

5. Background A service user’s perspective
There is a culture of sharing medications, so people may be taking something that was prescribed, but not for them.
Service users may only tell their psychiatrist what they think they want to hear; that is about medicines related to that speciality. Medication for physical health problems may not be mentioned. Oral contraceptives may be missed.
Also, herbal preparations, or medicines such as ibuprofen that can be bought without a prescription may not to be mentioned.
Mental health service users may be reluctant to tell a clinician that they have reduced or stopped taking their prescribed medication because they think they will be reprimanded.
There is a culture of people talking about what has worked for them and sometimes being generous enough to allow someone else a number of tablets so they can try it for themselves. So people may be taking something that was prescribed, but not for them.
When a service user is seeing a consultant for a physical problem as well as for a mental health problem, when asked about medication they may tell each consultant only what they think they want to hear, that is, the drugs to do with that specialty. This is because any medication for the other problem will be seen as not their concern (and it can seem as if the consultants see it like that too).
Also when someone is taking herbal preparations, especially if they have not been obtained from a Western medical source like a pharmacy or other shop (across-the-counter), they may not be mentioned when that person is talking about psychotropic medication they are taking. Ethnically derived herbal preparations, for example, will tend not to be mentioned.

6. Relevant Target Clinical practice
The aim of medicines reconciliation is that there is no unintentional discrepancy between medication prescribed prior to admission and medication prescribed at the point of admission
The audit data allow benchmarking of the proportion of patients for whom medicines reconciliation was not possible because fewer than two sources of information about their current medication were checked.
The target is a proxy measure of whether medicines reconciliation is likely to have taken place. It is not possible within the constraints of audit methodology to know whether the information given by the patient, carer, GP or specialist is 100% accurate and has been interpreted and acted on in a way that ensures medicines reconciliation has been achieved.
We use the proxy measure of evidence that at least 2 sources have been used to obtain information about medication the patient was taking prior to admission; in such cases medicines reconciliation was possible (although we cannot be sure it happened). Where no sources or only one source was checked, medicines reconciliation cannot have been achieved.

7. Method Benchmarking the quality of medicines reconciliation
Participating Trusts and clinical teams: acute adult, acute elderly and forensic wards (self-selected).
Audit inclusion criteria: patients recently admitted to ward and stayed for at least 7 days; minimum of 5 patients per ward advised.
Data collected for each patient:
Age, gender, ethnicity, diagnostic grouping, time of admission,
detainment under the Mental Health Act & ward type
Documented details of medicines prior to admission (prescribed
and non-prescribed), and adherence
Details of sources checked and discrepancies identified – from
discussion with the clinical team
Documentation of the reconciliation process
Time period of pharmacist and/or medicines management
technician involvement

8. Diagnosis not yet reached
National findings
Forty-three Trusts submitted data for 2296 patients from 455 clinical teams
Key demographic characteristics
Acute Adult N=1338
Acute Elderly N=683
Forensic N=275
Gender
Male
694
(52%)
274
(40%)
232
(84%)
Ethnicity
White/White British
980
(73%)
601
(88%)
182
(66%)
Black/Black British
128 (10%) 12
(2%) 52
(19%)
Asian
104
(7%) 14 11
(4%)
Mixed or other
126
(9%) 54
(8%) 30
(11%)
Age
Mean age in years (SD) 41
(13.6) 77
(8.9) 36
(13%)
Detained under Mental Health Act
Yes
658
(49%)
174
(25%)
249
(90%)
ICD-10
F00-F09 13
(1%)
269
(39%) 5
F10-F19 86
(6%) 18
(3%)
(5%)
F20-F29
555
(41%) 67
(10%)
168
(61%)
F30-F39
386
(29%)
224
(33%) 26
Other diagnoses
138 25 45
(16%)
Diagnosis not yet reached
160
(12%) 80
The majority of cases were admitted to either an acute adult or acute elderly ward. Age profiles and the proportion of patients detained under the Mental Health Act are largely in line with clinical expectations. The majority of the sample as a whole had a diagnosis of either schizophrenia spectrum disorder (F20-29) or affective disorder (F30-39), with the breakdown of diagnoses across clinical settings differing in line with clinical expectations.

9. National findings
Details about the medications prescribed prior to admission documented in the clinical records (n=2296)
This chart shows similar levels of documentation of the names and doses of medication in acute adult & acute elderly settings.

10. National findings
Documentation of adherence to prescribed medication in acute adult (n=1105), acute elderly (n=589) and forensic (n=239) services, for those in whom medication was prescribed
Documentation relating to adherence to medication regimens was more common in acute adult and forensic wards than elderly wards, which may reflect the assumption of health professionals that elderly patients are more likely to take their medication as prescribed. Another contributing factor could be that carers or residential home staff are available to confirm that medication has been taken.

11. Sources of information checked
Acute adult wards Acute elderly wards
Forensic wards
These three Figures show the sources checked by clinical teams within 24 hours, 3 days and 7 days of admission in each setting.
The figures show the bulk of sources are consulted within the first 24 hours of admission. With the exception of contacting GPs, there is little further evidence of reconciliation in the subsequent six days.

12. Discrepancies identified by each source within 24 hours, 3 days and 7 days
Acute adult wards Acute elderly wards
Forensic wards
These three Figures show the discrepancies identified by each sources checked by clinical teams within 24 hours, 3 days and 7 days of admission in each setting.
In these three graphs, for each source the total bar length represents the number of patients for whom the source was consulted (as a non-primary source). These graphs exclude consultation with primary sources, as by definition a discrepancy cannot be identified from the first source checked. The darker parts of the graphs (green, blue and purple) show the number of contacts in which a discrepancy was identified. The scale given is absolute patient numbers.
Overall, the GP was the source most likely to identify an unintentional discrepancy with the prescribed medication in adult and elderly wards. For forensic wards the source most likely to identify an unintentional discrepancy were non-forensic services. Asking the patient and examining the medication brought in also provided worthwhile information

13. Discrepancies identified
Proportion of discrepancies identified during medicines reconciliation by particular clinical team members
Discrepancies identified by:
Acute Adult
Acute Elderly
Forensic
Baseline
Re-audit
Pharmacist
168 (49%)
252 (48%)
80 (43%)
165 (57%)
16 (73%)
15 (71%)
Doctor
83 (24%)
75 (14%)
63 (34%)
38 (13%)
3 (14%)
Pharmacy/medicines management technician
43 (13%)
170 (32%)
28 (15%)
68 (24%)
0 (0%)
Other (including patient, nurse, ward clerk, or other member of team)
47 (14%)
31 (6%)
17 (9%)
17 (6%)
The data in the Table above suggest that in practice pharmacists and pharmacy staff take on the associated tasks.
The profile at re-audit differed to baseline for acute adult and acute elderly settings. A greater percentage of discrepancies were identified by members of the pharmacy team, most notably medicines management technicians. The forensic data are more difficult to interpret due to small numbers.

14. Clinical significance of the discrepancies identified
A small number of the discrepancies identified were clearly clinically significant
Omission of phenytoin, low molecular weight heparin and fentanyl patches
Wrongly transcribed lithium dose
Some had the potential to be clinically significant in the short/medium term
Omission inhalers for asthma/COPD, eye drops for glaucoma, antihypertensive medication
The clinical significance of the majority of discrepancies was difficult to determine
Changes in the time of administration of psychotropic medicines
Missing medicines for minor ailments (creams etc)
If your Trust opted to collect qualitative data relating to discrepancies identified during the process of medicines reconciliation, these have been returned to your POMH lead on a EXCEL file for further local analysis.

15. Trust level
Documentation in the clinical records of medications prescribed prior to admission for each Trust and in the total national sample (n=2296)
The Trust with the highest proportion of patients for whom both the name and dosage of medications prescribed prior to admission documented in the clinical notes is on the left hand side of the Figure and the Trust with the lowest proportion on the right. The proportions in the total national sample (TNS) are shown on the far right of the Figure.

16. Trust level
Documented information regarding patients’ adherence to prescription medication for each Trust and in the total national sample, for those in whom medication was prescribed
The Trust with the highest proportion of documentation regarding adherence in the clinical notes is on the left hand side of the Figure and the Trust with the lowest proportion on the right. The proportions in the total national sample are shown on the far right of the Figure. Practice at the baseline audit is shown by the white diamond. Where this is in the blue section of the bar, practice has improved since baseline.
The majority of the TNS was drawn from acute adult settings, and in this group the documented level of poor adherence is in line with the published literature.

17. Trust level
Subgroups of patients according to the number of sources checked within 7 days of admission.
For 485/2,296 patients (21%) in the red or orange sections (0 or 1 sources) reconciliation could not have taken place.
The Trust with the highest proportion of patients for whom two or more sources were checked is on the left hand side of the Figure and the Trust with the lowest proportion on the right. The proportions in the total national sample are shown on the far right of the Figure. For patients in the red or orange sections (0 or 1 sources) reconciliation could not have taken place, as by definition there must be two or more sources checked for this to be possible.
*Trust 56 did not submit sufficient data to allow data analysis at re-audit.

18. Discrepancies identified
The proportion of patients in each Trust for whom two or more sources were checked (i.e. medicines reconciliation was possible), and the proportion for whom discrepancies were identified (n=1811)
The Trust with the highest proportion of patients for whom two or more sources were checked is on the left hand side of the Figure and the Trust with the lowest proportion on the right. The proportions in the total national sample are shown on the far right of the Figure.
*Trust 56 did not submit sufficient data to allow data analysis at re-audit.

19. Key findings: Clinical Practice
RELEVANT TARGET
For 21% of patients in the total national sample, fewer that 2 sources of information about currently prescribed medication were checked (range across participating Trusts from 0% to 82%). Thus, medicines reconciliation would not have been possible in these cases.
During the process of medicines reconciliation, discrepancies were identified in 25% of the total national sample for whom two or more sources of information had been checked.
Of the specific discrepancies reported, only a small proportion were clearly clinically significant.

20. Key findings: Clinical Practice
Proportion of patients in whom at least 2 sources were checked (reconciliation was possible) at baseline and re-audit in your Trust and the total national sample (TNS).
Baseline
Re-audit

21. What happens next?
Your Trust’s Local Project Team Lead (LPTL) has a copy of the full report, please ask them for a copy to see these findings in more detail.
Slide set containing key findings from the re-audit.
Examples of good practice will be shared on the POMH-UK website.
No current plans to re-audit.
If you have any questions, please ask your LPTL or POMH