Obstetric Anaesthetists’ Association/National Perinatal Epidemiology Unit collaborative project to develop key indicators for quality of care in obstetric anaesthesia: first steps in the right direction
The aim of delivering high-quality clinical care that optimises patient outcomes has always been a driving principle in the practice of medicine. However, how best to measure and assess the quality of care provided has only become a focus in the past few decades. Quality indicators to assess healthcare structures, clinical processes, patient experience and outcomes related to the provision of anaesthesia are now a priority for many institutions in the USA (e.g. American Society of Anesthesiologists, Anesthesia Quality Institute, Agency for Healthcare Research and Quality, Multicenter Peri-operative Outcomes Group) and the UK (e.g. Department of Health, National Institute for Health and Care Excellence). In recent years, there has been an exponential growth in studies evaluating quality and outcome measures; a 2018 systematic review of studies published in the past 10 years identified 43,860 journal articles and 43 relevant indicator programme publications, with a total of 1282 clinical indicators described 1.
The development of indicators to measure and assess the quality of obstetric anaesthetic care, and determine if the quality of care or processes are linked to improved outcomes, is currently very limited. In this edition of Anaesthesia, Bamber et al. report the findings from an Obstetric Anaesthetists’ Association (OAA) and National Perinatal Epidemiology Unit (NPEU) collaborative that aimed to identify a core set of quality indicators for obstetric anaesthetic services 2. This collaborative effort consisted of a three-stage Delphi survey of 133 key stakeholders that included obstetric anaesthetists, other maternity care health professionals and women who had used maternity services. The initial 31 indicators included for evaluation covered three domains of service provision (13 indicators); service quality (8 indicators); and clinical outcome (10 indicators). These were derived from national peer-reviewed publications, with the option for stakeholders to make suggestions for additional indicators. These quality and outcome indicators were pared down by the staged Delphi process and final consensus to five indicators that the stakeholders considered the most relevant to the assessment and benchmarking of the quality of care in obstetric anaesthesia 2.
- Proportion of women who had an epidural (or combined spinal-epidural) for labour analgesia who had an accidental dural puncture.
- Availability of guidelines for the referral of patients to an anaesthetist for an antenatal review.
- Existence of elective caesarean section lists with dedicated (i.e. not expected to cover emergency work) obstetric, anaesthetic and theatre staff.
- Availability of point-of-care testing for estimation of haemoglobin.
- Proportion of epidurals for labour analgesia that provided adequate pain relief within 45 min of placement.
Two of the proposed indicators, ‘Whether there is O rhesus-negative blood immediately (within five minutes) available at all times for emergency use’ and ‘Whether there is at least one fully equipped and fully staffed obstetric theatre within the delivery unit’ were excluded from the core set, as the consensus was that they were already attained by nearly all units in the UK 2.
The OAA/NPEU collaborative needs to be congratulated on this excellent attempt to define key quality indicators for obstetric anaesthesia care. The authors propose that these quality metrics can be used to support the hospital's quality improvement activities and help in the development of national benchmarking quality standards in obstetric anaesthesia. It is, however, important to reflect further on these quality indicators before widespread adoption is considered. The initial list of quality indicators considered the key domains of service provision, service quality and clinical outcome, but few items considered were patient-centred metrics. This is not surprising, as only 10% of the stakeholders were service users, and therefore the final indicators should be considered physician- and not patient-derived. Patient-reported outcomes and experience of care measures are important indicators to consider, and a more patient-centred approach is necessary to determine quality metrics considered important for women who had used or will use maternity services. Obstetric anaesthetists made up 71% of the initial stakeholders. Different stakeholders in healthcare may have very different perspectives on the importance of healthcare quality indicators 3. Greater input from obstetric shareholders (obstetricians, midwives and nurses) is essential for indicators to be balanced and less anaesthesia-provider focused.
A high proportion of the quality indicators in the final phase and core set were process indicators, and outcome measures were very limited. The provision of optimal care during childbirth is important to improve maternal mortality and morbidity. The Centers for Disease Control and Prevention reported that 60% of deaths during childbirth were preventable, and the California Maternal Quality Care Collaborative determined that facility factors contributed to 75% of preventable fatal outcomes 4. The bias toward service provision compared with service quality or outcome may reflect the anaesthetist-heavy stakeholder composition; lack of consensus for optimal obstetric anaesthetic outcomes that are routinely collected or considered; and the inherent rarity of maternal mortality or morbidity. Additionally, the primary consideration at the consensus meeting was the measurability of the quality indicators, and as such, process indicators predominating in the final core is not surprising. This shortcoming related to quality indicators, outlined above, is not unique; a systematic review of published studies of quality indicators in anaesthesia found that 64% were process indicators, and that patient-centred metrics accounted for the fewest published clinical indicators 1. Objective data are critical for any quality improvement initiative, and while the OAA/NPEU collaborative considered process and outcome metrics, balancing metrics need to be developed to fully determine the value of an intervention. Balancing metrics should also be introduced to measure and ensure there are no unintended negative consequences of any quality initiatives, for example, the percentage of epidurals for labour analgesia that provided adequate pain relief within 45 min of placement metric may lead to hurried techniques or non-titrated dosing practices which may compromise safety, or the indicator may lead to more combined spinal-epidurals vs. standard epidurals. Lastly, a causal link between indicators and improved outcomes or increased maternal safety needs to be demonstrated in subsequent evaluations. However, this link is likely to be challenging to prove, and most proposed anaesthetic quality indicators do not have a level of evidence ascribed to them in the literature 1. However, the introduction of quality indicators may indirectly improve care, as implementation generally requires interdisciplinary discussions and associated institutional behavioural changes.
The Delphi method is now widely used to form expert consensus on healthcare questions, and the OAA/NPEU collaborative's criterion that quality indicators needed to be rated as extremely important by > 90% of stakeholders in at least two panels was well intentioned; however, trying to produce just a handful of quality indicators that encompasses the broad scope of practice for obstetric anaesthesia is very difficult. An unintended consequence of distilling quality of care down to just a few metrics is that other key care elements may be de-emphasised. An optimal model of quality assessment should consider an organisation's structure (how care is organised), processes of care (what is done) and healthcare outcomes (the achieved results) 5. Structure indicators should include physical resources (facilities and equipment) and human resources (number, qualifications and availability of personnel), and although the OAA/NPEU collaborative did consider several structure indicators, the final metrics cannot possibly encompass this adequately. For example, point-of-care testing for estimation of haemoglobin availability is of limited value unless a massive transfusion protocol and emergency release system for available blood, a blood bank protocol that is tested and functional and a rapid-infuser device to assist with massive resuscitation, are readily available for use in the obstetric unit. A different study methodology and collaborative composition may have resulted in different metrics being obtained. To try to provide a broader evaluation of the quality of obstetric anaesthetic care, the Society of Obstetric Anesthesia and Perinatology (SOAP) created a programme to designate Centers of Excellence for obstetric anaesthetic care 6. This approach moves the focus away from creating a select few quality indicators, towards a broad assessment of many domains including personnel and staffing; equipment, protocols and policies; simulation and team training; obstetric emergency management; caesarean section and labour analgesia care; recommendations and guidelines implementation; and quality assurance and patient follow-up systems 7. However, this approach requires active application and peer-review assessment and does not facilitate national benchmarking of quality standards for centres that choose not to apply or obtain designation.
Lastly, before broader application of these quality metrics, some indicators require clearer wording. Indicator number 3 asks ‘Whether there are elective caesarean section lists with dedicated (i.e. not expected to cover emergency work) obstetric, anaesthetic and theatre staff.’ In many units, there may be occasions where the delivery suite becomes very busy, with an impact on the flow of the elective caesarean section list, but how often is this allowable before the bracketed clause is deemed to be breached? Indicator number 5 requires the proportion of epidurals for labour analgesia that provided adequate pain relief within 45 min of placement (from the start of epidural insertion). However, what is adequate pain relief: no pain; pain score < 3; pain at a satisfactory level for the woman; etc.? When is the start of the epidural insertion: time the anaesthetist entered the room; started washing his/her hands; opened the epidural pack; inserted the first needle into the skin? It is crucial that all units are using exactly the same definitions for any meaningful comparisons to be made.
In conclusion, the authors have made a very important first step towards the creation of quality indicators for obstetric anaesthesia care in the UK. However, before these quality indicators are broadly applied, they require some further refinement. Subsequent stages in the process might be to provide some more patient-centred input, consider balancing metrics and initiate broader discussion on how to measure any impact on maternal outcomes following the introduction of these indicators.
Acknowledgements
SK is an Editor of Anaesthesia. No other competing interests declared.
