Anaesthetic records: an evaluation of variation across England

Summary

We collected blank non‐specialist anaesthetic records from 71 National Health Service Trusts in England. A data set was established by collating all data items found in an initial tranche of 28 records. All 71 records were subsequently analysed for each data item in this data set. We found significant variation: the most populated record included 216 data items and the least included 38 data items: a greater than five‐fold variation. There was significant variation in the inclusion of data items commonly considered important to patient safety; 42% of records omitted documentation of fasting status, 72% omitted documentation of a discussion around the risk of accidental awareness during general anaesthesia, 92% omitted documentation of quantitative neuromuscular blockade monitoring and 63% omitted documentation for ‘Stop Before You Block’ when performing regional anaesthesia. The study highlights significant variability in the composition of anaesthetic records across England which may impact on its value as a data repository, an action trigger, a medicolegal account, and a tool to facilitate safe handover. Standardisation of the anaesthetic record or the establishment of standards of recording would help to allay potential risks to patient safety and assist in guiding future procurement of electronic solutions for anaesthetic records.

Introduction

An anaesthetic chart (or record) is used to log the conduct and chronology of anaesthetic care and, whether in paper or electronic form, is mandated as part of the Guidelines for the Provision of Anaesthetic Services (GPAS) of the Royal College of Anaesthetists (RCoA) [1]. It performs multiple functions, acting as data repository, action trigger, medicolegal account, and as a tool to facilitate safe handover. Locally developed anaesthetic record templates could result in significant variation in documentation practices. Standardisation is an effective mechanism for reducing human error in complex situations. Examples include ‘Situation, Background, Assessment, Response’ (SBAR) handover [2], the use of 2222 as the European cardiac arrest call number [3], and drug syringe labelling colours [4]. The getting it right first time (GiRFT) programme emphasises reducing ‘unwarranted variation’, ‘learning from top‐performers’ and ‘benchmarking’ as key strategies for improving healthcare, and has been rolled out across hospitals in the UK [5]. Identification of variation in anaesthetic record design, benchmarking and subsequent standardisation has potential benefits which might include improving patient safety.

The Association of Anaesthetists makes detailed recommendations for standards of monitoring during anaesthesia [6], but almost no reference to standards for documentation. It advises that anaesthetic records are structured so that commonly monitored variables are recorded every 5 min, with others at 15‐min intervals [5]. Elsewhere, a guideline for effective documentation in regional blocks is being developed [7].

In conjunction with The Royal College of Anaesthetists (RCoA) and The Professional Record Standards Body (PRSB), a project exploring the possibility of setting minimum standards for non‐specialised anaesthetic records began in 2020 [8]. Here, we report an element of this project which aimed to collect and review the content of available anaesthetic record templates in use in England.

Methods

An initial sample of anaesthetic records was requested through the RCoA Clinical Leaders in Anaesthesia Network in August 2020. The 28 records received were analysed for the presence of similar sections, identifying five consistently present domains (Box 1) from which data items were extracted to create a full dataset (Supporting Information, Appendix S1).

Box 1. Consistent domains of anaesthetic records.

Domain 1 – pre‐operative assessment: 81 potential data items.

Domain 2 – pre‐operative discussion: 39 potential data items

Domain 3 – drug administration, physiological observations and events: 49 potential data items

Domain 4 – anaesthetic technique: 119 potential data items

Domain 5 – post‐operative handover: 12 potential data items

Using the NHS A‐Z of NHS Trusts in England [9], we contacted anaesthetic departments in September 2022 and again in October 2022 where necessary, requesting an anaesthetic record template from a consultant anaesthetist identified as a point of contact.

Each record was examined for the presence of each potential item found in the dataset from the original 28 templates. The proportion of included data points for each domain was calculated. This process was repeated to verify accuracy, with no errors identified in the second data extraction.

Ethical committee approval was not sought for this service evaluation as no patient data was accessed.

Results

Of 215 NHS Trusts in England, 81 were ambulance, mental health or paediatric trusts, leaving 134 who might contribute an anaesthetic record to the project. All these Trusts were invited, and 71 (53%) Trusts agreed to participate, including returns from every School of Anaesthesia in England ranging from 13 Trusts (London) to 3 Trusts (Wessex).

All records included data items in each of the five domains with at least 75% of records including at least 70 key data items. The data items included in at least 75% of records are presented in Table 1. Data by domain is shown in Supporting Information, Tables S1–S5 (Appendix S2). Results below are presented as absolute numbers (median, [interquartile range]) or incidence (proportion, %).

Table 1.

The 70 data items included in greater than or equal to 75% of records.

Item	% Inclusion
Domain 1
Patient name OR ‘attach addressograph here’	100
Trust	100
Assessing anaesthetist	100
Date of assessment	100
ASA grade	97
Allergy – causative agent	97
Weight	92
Clinical summary/past medical history	90
Proposed operation	86
Date of birth	85
Hospital number	82
Platelet count	82
Serum sodium	82
Serum potassium	82
Serum creatinine	82
Anaesthetic history	80
Height	79
Body mass index	77
Blood pressure	77
Serum urea	77
ECG	76
Domain 2
Anaesthetic agreed based on risk/benefit	83
Neuraxial anaesthesia – procedure explained	80
General anaesthesia – procedure explained	79
Regional anaesthesia – procedure explained	77
Domain 3
Observation grid (graphical)	100
Blood pressure (visual)	100
Heart rate (visual)	100
Medication grid	99
Dose	99
Observation grid (numerical)	99
End‐tidal carbon dioxide (numerical)	92
Free text	92
End‐tidal anaesthetic agent (numerical)	89
Fraction of inspired oxygen (numerical)	87
Intravenous fluids (numerical)	87
Oxygen saturations (numerical)	83
Temperature (numerical)	79
Urine output (numerical)	77
Reference to neuromuscular monitoring	75
Domain 4
End‐tidal anaesthetic agent monitoring	97
End‐tidal carbon dioxide monitoring	97
Temperature monitoring	97
Anaesthetic technique – other	96
Laryngoscopy – grade	96
Non‐invasive blood pressure monitoring	94
Anaesthetist 1	93
Endotracheal tube – size	93
Fraction of inspired oxygen monitoring	93
Other monitoring	91
Date of anaesthetic	90
ECG monitoring	90
Eye protection	90
Supraglottic airway – size	90
Warming blanket	89
Invasive blood pressure monitoring	87
Other patient safety	86
Peripheral line 1 – site	86
Warmed fluids	84
Machine/circuit check	83
Supervising anaesthetic consultant	83
Patient position for surgery	81
Urgency	81
Operating surgeon	80
Peripheral line 1 – gauge	80
Intermittent positive pressure ventilation	77
Neuromuscular monitoring	77
Ventilation – spontaneous	77
Breathing system	76
Pressure areas protected	76

ASA, American Society of Anesthesiologists; ECG, electrocardiogram.

Domain 1: pre‐operative assessment

Of 81 potential data items, the individual records reviewed included between 5 and 51 items (35, [29–42]). This domain covered the smallest area in all records.

Domain 2: pre‐operative discussion

Of 39 potential data items, the individual records reviewed included between 1 and 30 items (16, [8–18]).

Domain 3: drug administration, physiological observations, and events

Of 49 potential data items, the individual records reviewed included between 9 and 32 items (22, [18–25.5]). Records included space to record the administration of a median of 11 medications (IQR 9–12) and observations lasting a median of 5 h (IQR 4–6). The lowest total number of spaces for medications was seven (three records) and the shortest recordable duration of anaesthesia was 2.5 h (two records). This domain covered the largest area in all records.

Domain 4: anaesthetic technique

Of 119 potential data items, the individual records reviewed included between 11 and 114 items (68, [55–80]). Regarding regional anaesthesia, 18 (26%) of records specified ‘block laterality’ and 26 (37%) triggered documentation of ‘Stop Before you Block' or similar [10].

Domain 5: post‐operative handover

Of 12 potential data items, the individual records reviewed included between 1 and 12 items (5, [2–7]). In two records (5.5%), the domain was a blank space with no prompts and there was one prompt in a further two records (5.5%).

All domains

From 300 potential data items, the individual records reviewed included between 38 and 216 items (143, [120.5–161]) equivalent to 13%–72% inclusion (Fig. 1).

Figure 1.

Percentage of potential data items included in anaesthetic records in English NHS Trusts (Trusts anonymised for publication purposes).

Illustrative anaesthetic records are shown in Figure 2a and b.

Figure 2.

(a) The most highly populated anaesthetic record; (b) the most sparsely populated anaesthetic record.

Discussion

We believe this is the first evaluation of variability in anaesthetic records on a national level. The results demonstrate wide variation in anaesthetic record templates, with a 5.7‐fold variation in the number of data items included. Variation exists across all five documentation domains, including within areas where guidelines and standards for anaesthetic care exist.

While not all variation in healthcare is detrimental, unwarranted healthcare variation is generally considered undesirable. It may be associated with variation in quality of care and can usefully generate exploration of whether reduced variation is desirable [5]. While addressing unwarranted variation may progress to the identification of good (best) practice, benchmarking and standardisation, the first step is an examination of variation. Our examination of anaesthetic records from most English Trusts shows high levels of variation in design and content that might be considered concerning from a patient safety, medicolegal and research perspective. It therefore provides an opportunity to discuss what is optimal content and for future benchmarking and standardisation.

With the many functions of the anaesthetic record described above, it is plausible that substantial variation might be associated with or even cause variations in care delivery. Some anaesthetic records are very sparse and rely considerably on the anaesthetist to recall and self‐report all the necessary data fields. This approach is prone to unreliability which may impair history taking, documentation of pre‐operative discussions, accuracy or omission of information relevant to risk–benefit discussions, the quality of shared decision‐making, and the future ability to discover when and why a patient has previously been challenging to anaesthetise. Conversely, it may be that not all 300 items identified from the review should be included for reasons of usability. A minimum or ideal dataset should usefully be agreed to improve consistency.

While a complete discussion of the results is beyond the scope of this article, we describe several illustrative areas of interest.

Difficulty in airway management significantly impacts anaesthesia‐related morbidity, mortality, and litigation [11, 12]. The 4th National Audit Project of the RCoA and Difficult Airway Society reported that poor airway assessment contributed to poor outcomes [13] and airway incidents may be both life‐threatening and medicolegally expensive [12]. Airway assessment is fundamental to safe anaesthesia and clear, consistent documentation indicates it was undertaken, records its results and can link to a logical airway plan. These data may prove invaluable in the event of airway difficulty in future anaesthetics [14]. Such documentation is therefore important for the index anaesthetist, for future anaesthetists, for medicolegal reasons and for patient safety.

A well‐designed anaesthetic record can act as an action trigger for asking critical questions and documentation of pertinent data. Almost half the records had no prompt for pre‐operative airway or neck mobility examination, and little more than half prompted questions about starvation status or symptoms that increase aspiration risk. While their absence in a standard template does not prevent the anaesthetist from documenting all necessary information impeccably, it relies on the ‘human as hero’ which is an unreliable way of ensuring safe care [15]. If it is logical that anaesthetic records include data items to document a physical airway assessment, aspiration risk and an airway strategy, it follows that a minimum standard for such documentation should be agreed. Similar arguments apply to many of the data items variably present across the examined anaesthetic records including patient assessment, communication of process and risk, collection of physiological data or handover of care to other professionals.

It is established that pre‐operative patient information should be provided and shared decision‐making undertaken before the day of anaesthesia. Despite this, the use of a separate ‘anaesthesia consent from’ is not part of normal UK anaesthetic practice and anaesthetists commonly discuss treatment options, risks and benefits in the pre‐operative clinic or at the immediate pre‐operative consultation. Many anaesthetic records include a section to facilitate documentation of these discussions which is important routinely and medicolegally, particularly as the landmark Montgomery case now necessitates discussion of ‘material risks’ [16]. Lack of time is the most common barrier to shared‐decision making cited by clinicians [17]. Well‐structured record templates for common risk discussion points could aid the efficient delivery and documentation of information provision and shared decision‐making. This is supported by recent Delphi consensus around documentation of patient discussions before regional anaesthesia [6]. While a tick box list of topics discussed may act as an aide memoire or minimum requirement (but not a maximum requirement), a ticked box is no more a measure of the quality of consent than is a signature on a form. Tick‐boxes may act as a reminder to the clinician that a discussion took place but are no substitute for a personalised and complete exploration of matters of importance to the patient.

The Joint Commission cited inadequate staff communication during patient handover as the leading cause of ‘sentinel events’ in 2022 [18]. Studies implementing handover checklists show a reduction in adverse events and overall medical error [19, 20]. Despite this, the postoperative instruction section of the analysed charts was sometimes a blank or nearly blank space. Standardisation of this aspect of anaesthetic documentation presents an opportunity to improve anaesthetists' communication practices at this vital care transition point.

The heterogeneity of data collected by anaesthetic records nationally also presents a barrier to effectively sharing data for research, audit, and transfer of care. Standardisation could also support anaesthetists in training who rotate from one hospital to another, providing them with increased consistency. Standardisation could assist in medicolegal settings by creating a benchmark for documentation and avoiding uncertainty about the adequacy of any individual anaesthetic record template.

Lastly, electronic anaesthetic information management systems (e‐AIMS) are increasingly used and could provide benefits for patients, clinicians and managers beyond the simple automation of the paper record. Most anaesthetic activities do not have specific speciality codes due to the limited lexicons for anaesthesia provided by the International Classification of Disease (ICD‐10) and Office of Population Censuses and Surveys (OPCS‐4), with the result that much of the work done by anaesthetists in and out of theatre is invisible [21]. This has implications for workforce planning, perioperative team development and departmental finances. The future design and implementation of e‐AIMS would greatly benefit from a national standard for anaesthetic records.

There are some limitations to this report. We have studied little more than half of England's anaesthetic records. There is no standard against which we can judge and therefore we have simply described what we have found on each record. Some aspects of documentation may be dealt with in adjunctive documents such as local safety standards for invasive procedures which we have not captured.

In conclusion, this project highlights significant variability in anaesthetic records across England. The variability is a concern and is unlikely to be clinically justified or represent and support best practice. An ideal anaesthetic record would collect a sufficient dataset for documenting pre‐operative discussions and decisions, explain the narrative of the case, populate hospital databases, support audit/clinician handover and provide its clinical trigger role: all without its completion being burdensome or distracting from clinical care. A very sparse document may cognitively overload the person completing it (omissions are likely) and well‐organised tick‐box designs appear easy to use. Tick‐boxes do not guarantee completion of all aspects of the complex process of anaesthesia but nonetheless act as a useful prompt. The speciality, through research‐based consensus or professional working party, could usefully agree which data points an anaesthetic record should contain. This could then be benchmarked and agreed standardisation implemented. The increasing adoption of e‐AIMS strengthens the need for such agreement.

Supporting information

Appendix S1. Full dataset with descriptors.

Appendix S2. Supplementary data tables.

Table S1. Percentage inclusion of data items in pre‐operative assessment.

Table S2. Percentage inclusion of data items in pre‐operative discussion.

Table S3. Percentage inclusion of data items in drug administration, physiological observations and events.

Table S4. Percentage inclusion of data items in anaesthetic technique.

Table S5. Percentage inclusion of data items in post‐operative handover.