Composite indicators are seen as an objective and efficient method of measuring performance in the NHS. In this month's column, I will examine both the strengths and the potential pitfalls of this method of assessment in relation to the provision of UK maternity services. Contained within the discussion of what statistical data to gather, as well as the choice of format to present it in, are subtler political and economic outlooks that have a tendency to confer importance on empirically measurable variables. Within this construct we encounter the potential danger of devaluing other equally important, but less quantifiable, aspects of midwifery practice. In addition, the subsequent assembling of national league tables, with insufficient examination of the differing socioeconomic demographics involved, causes me to question a number of aspects of the exercise in its current form.
So what are they?
In their broadest sense, composite indicators can be regarded as an index or integration of individual (but related) performance indicators. This index is often presented in a tabulated format alongside other such indicators and aims to represent the distillation of a large amount of information in a manner that is comprehensible and widely distributable. Composite indicators are used internationally for a variety of purposes. In terms of the NHS, they embody an established tool for governance and efficacy comparison. They also indirectly link to:
‘Composite indicators are globally legitimised, powerful comparative tools with which to highlight areas of healthcare service provision’
They have informed the new NHS Commissioning Board with regard to their future incorporation into the clinical reference group (CRG) for the women and child programme (NHS Commissioning Board, 2013). It is not within the remit of this opinion column to provide an extensive critique of composite indicators (for a concise summary see Jacobs et al, 2004). Nevertheless, because of their ubiquitous expansion into various spheres, it is useful to briefly explore their role in healthcare and maternity services.
Why do we have them?
Composite indicators have evolved because they are predicated on a variety of rationales that are principally directed towards promoting accountability by controlling and setting standards for healthcare system provision (Smith, 2002). In this sense, they underpin efficiency agendas and prioritise the performance of a service towards specific indicators assigned for measurement. If what is measured encompasses the leading features of any given department, composite indicators are deemed to be effectively representative. They enable a balanced judgment concerning an organisation's activities and facilitate decisions concerning improvements that may be required. Composite indicators also allow comparison to be made against an ‘equivalent’ service. This has lead to their use in performance league tables that contrast healthcare institutions at both national and international levels. Thus composite indicators are globally legitimised, powerful comparative tools with which to highlight areas of healthcare service provision that might range from simply undesirable to potentially dangerous. For example, by drawing attention to discrepancies in mortality rates between NHS Trusts and health practitioner working patterns (Dr Foster Intelligence, 2014).
So what is wrong with them?
Composite indicators have been critiqued for several reasons. Perhaps more obviously, the separate performance indicators selected for inclusion in the composite indicators often invoke contention. It is also suggested they distort practitioner behaviour. The increase in antenatal screening options, arising over the past decade, possibly illustrates this phenomenon. Arguably, antenatal screening is now a prime remit of early midwifery care that is subjected to such monitoring. Of course influencing practice is extremely welcome when it highlights and addresses key issues such as evidence-based interventions. However, this performance-driven focus on providing screening information may be detrimental to situations requiring subtler aspects of communication with women, such as exploring mental health matters or fears about birthing.
Critics also suggest that the mechanisms for producing composite indicators can be complex because they rely on the ‘weighting’ attached to particular indicators. In addition, they do not always comprehensively adjust for variables such as sociodemographics (Freudenberg, 2003; Jacobs et al, 2004). These issues may confer composite indicators with the status of being a ‘blunt tool’ for service efficacy comparison. Moreover, in common with certain areas of statistical acquisition, the data composite indicators rely on may be insufficiently gathered for particular indicators. This leads to legitimate questions concerning the accuracy and veracity of overall results. For example, I recall the conundrum I faced in some situations when attempting to classify the first feed of an infant. This became problematic when a mother appeared conflicted regarding her feeding intentions. When she offered her baby no more than a minimal amount of time at the breast, I wondered whether that ‘feed’ counted as initiation of breastfeeding or not?
Where do they come from?
Having considered why the use of composite indicators has become so ubiquitous in healthcare service provision, questions arise concerning the origins of their current form. Their concerted use in England was instigated by New Labour in 2000 when the Star ratings system for NHS Trusts was introduced (Klein, 2010). Composite indicators were implemented as an assessment tool for Trusts and were linked to fiscal ‘rewards’ for policy compliance and ‘fines’ for underperformance. This free market approach to health care provision, embraced wholeheartedly by the government at the time, also placed increased emphasis upon the concept of patient choice. It was envisaged that composite indicators would generate ‘self-correcting’ behaviour in a Trust's operational performance. From this perspective, Trusts were thought to be highly motivated to optimise outcomes of their composite indicators because otherwise the public might be disinclined to ‘choose’ healthcare services that were ‘demonstrably poor’. One flaw in this policy seemed obvious to me a decade ago, when I questioned just how accessible neighbouring Trusts' maternity services could be for pregnant women? Research also presents a fascinating insight into how composite indicator indexes and league tables appear to the public and health professionals. Despite people generally welcoming composite indicators, several studies suggest the information they contain is viewed as untrustworthy. Even mortality statistics are not as influential on healthcare consumer choices as the views of their peers. However, for the demographic that includes educated, younger professional people there is some evidence to the contrary (Marshall et al, 2002).
‘Composite indicators were implemented as an assessment tool for Trusts and were linked to fiscal ‘rewards’ for policy compliance and ‘fines’ for underperformance’
Where are they going?
In terms of current maternity services, there are a variety of mechanisms that provide material for what could be regarded as composite indicator purposes. These include, as previously mentioned, statistics that may be gathered from healthcare practitioners for the purposes of CRMS, KPIs, HES and CRGs. Yet increasingly information is being obtained directly from women using maternity services. This is happening through an assortment of quality assurance mechanisms that may be instigated by: NHS Trusts, patient interest groups or the healthcare service inspectorate. For example, the Care Quality Commission (CQC, 2014) are proposing some interesting new indicator criteria for their next maternity services survey due in 2016. These include specific questions relating to how both women and their partner's worries or concerns were addressed by maternity staff, and whether people felt they were treated with dignity and respect.
Despite this planned development by the CQC, the current governance agenda predominantly formulates contemporary composite indicators using an evidence-based, clinically focused direction. I fully acknowledge that this approach has played a central role in improving care for women and maternity service standards. Yet they could be viewed as somewhat reductionist and lacking in holism. They pay little, if any deference to the emotional and psychological measurement of the consequences of maternity care. Of course a clinical focus is critical but as the Royal College of Midwives (RCM), Royal College of Obstetricians and Gynaecologists (RCOG) and the National Childbirth Trust (NCT) (2013:1) jointly state ‘maternity care has a profound impact on women's physical, emotional and psychological health throughout their life’. With suicide rates representing a highly significant statistic in maternal mortality, it could be argued that the current priorities of composite indicators in maternity services may require revision to further encompass these emotional and psychological indicators.