Perceived Inappropriateness of Care in the ICU: Title and subTitle BreakWhat to Make of the Clinician’s Perspective?

The patient's perspective is now rightly accorded a position of great importance in determining the quality of health care delivery. In the intensive care unit (ICU), for example, clinicians are no longer interested only in ensuring that patients survive to hospital discharge or avoid ventilator-associated pneumonia, but also in achieving long-term, quality-adjusted survival¹ for patients and avoiding posttraumatic stress among surrogate decision makers.² Yet the frontline clinicians who provide critical care have rarely been considered to be relevant stakeholders in determining whether the care they deliver is consistent with the care they should deliver.

The APPROPRICUS study, reported by Piers and colleagues in this issue of JAMA,³ suggests that a new, more inclusive approach may be in order. Piers et al³ examined perceived inappropriateness of care among 1651 ICU physicians and nurses providing care on a single day in 82 adult ICUs across Europe and Israel. Defining perceived inappropriateness of care as care that “clashes with [the clincian’s] personal beliefs and/or professional knowledge,” the authors found that more than one-fourth of clinicians believed they were providing inappropriate care during this 1 day alone. Roughly two-thirds of these clinicians reported providing care that was disproportionate (typically too intense) to what they believed was warranted for that patient, and one-third reported providing ICU care to 1 patient that they believed could better serve another.

The APPROPRICUS study yielded several additional findings of note. Perceived inappropriateness of care was found to be associated with clinicians' reported intentions to leave their jobs, raising concerns about the sustainability of the workforce necessary to provide critical care if this problem is not remedied. Clinicians practicing in ICUs rated as having suboptimal collaborative environments were more likely to report perceived inappropriateness of care, raising the possibility that center-level characteristics might be targets for interventions designed to reduce perceived inappropriateness of care. In addition, high levels of discordance were observed between clinicians (eg, attending physicians and nurses) in their views of the appropriateness of care for the same patient on the same day. The authors note that such discordance highlights the subjective nature of perceived inappropriateness of care and its potential to serve as a marker of poor communication, decision sharing, or job autonomy in the ICU. However, the same observation could instead signal a lack of construct validity of the measure, raising questions about how common or consequential perceived inappropriateness of care really is.

Overall, the APPROPRICUS study builds from smaller studies^{4 - 5} to provide the best evidence to date regarding the epidemiology of perceived inappropriateness of care. Generalizing the results beyond European and Israeli ICUs is challenging because most of the ICUs studied used closed staffing models, low patient-to-intensivist ratios, and 24-hour attending intensivist supervision—charactertistics not shared by many US or Canadian ICUs. Nonetheless, the large and diverse sample, remarkably high response rate, strong conceptual model, and ability to link clinician observations to individual patients are all substantial strengths. Furthermore, findings such as discordant perceptions of inappropriate care may in fact be higher in the United States, where multiple attending physicians and consultants may be involved in care decisions. Thus, it is important to ask why such a large number of clinicians are providing care they perceive to be inappropriate, and what can or should be done about it?

If so many clinicians are providing care that is not motivated by its appropriateness, what then are they trying to accomplish? Are clinicians heeding the desires of surrogates against their own judgment or attempting to minimize disagreement among members of the clinical team? APPROPRICUS suggests that at least some ICU clinicians believe they are acting appropriately when they responsibly steward scarce resources, but future work will be needed to determine what else motivates ICU clinicians' provision of care. Even if other goals, such as promoting family members' interests, are more important than clinicians' vantage points, it is concerning that so many clinicians are providing care they perceive to be inappropriate.

Ultimately, the specific implications of this study hinge on whether perceived inappropriateness of care is a valid and important outcome measure. Conceptually, providing care thought to be inappropriate could have important consequences for patients today, patients tomorrow, or clinicians themselves. First, perceived inappropriateness of care could engender retribution toward current patients or their families, degrading professional ethics and reducing the quality of care such patients receive. This possibility is supported by considerable face validity, but to date, no evidence.

Second, if repeated episodes of perceived inappropriateness of care increase clinicians' moral distress, then professionalism may erode over time or clinicians may leave the workforce altogether, both of which threaten the quality of care for future patients.⁶ These possibilities are supported by correlational but not causal data. For example, the study by Piers et al³ corroborates prior evidence that perceived inappropriateness of care is associated with burnout syndrome,^{7 - 8} which is in turn associated with reduced productivity, reduced ratings of job performance, and greater intentions to leave one's present job.^{9 - 10} However, none of these associations have been shown to meet traditional criteria for causality.¹¹ No temporal or dose-response relationships have been identified, and the strengths of the associations generally have been low. Perhaps most importantly, the bulk of data in this nascent field document associations between 2 concomitantly measured subjective end points, enabling several alternative explanations for the results. Does the provision of perceived inappropriateness of care cause clinicians to intend to leave their jobs or do otherwise disgruntled clinicians more commonly report both perceived inappropriateness of care and intentions to leave their jobs?

Third, if perceived inappropriateness of care represents a true threat to clinicians' moral integrity,¹² then reducing perceived inappropriateness of care may be important in its own right, without necessarily considering consequences for patients. Upholding clinicians' moral integrity or the preservation of their core moral beliefs requires that clinicians not be forced to provide care that they find so troubling that it would represent an act of self-betrayal and lead directly to moral harm.¹³ This reasoning underlies arguments suggesting that physicians may make conscience-based refusals or conscientious objections to providing many forms of care. Such claims typically have been made to prevent clinicians from providing care they perceive to be wrong for the individual patient, such as those cited in two-thirds of the perceived inappropriateness of care cases in APPROPRICUS. However, conscientious objections conceivably could be based instead on concerns about distributive justice (cited in one-third of cases), if the clinicians believed that violating distributive justice was so wrong that doing so would betray their respective core moral identities.

There is as yet no consensus on what, if any, normative weight ought to be placed on clinicians' moral integrity as a factor governing quality in health care. Indeed, serious challenges have been levied against efforts to incorporate clinicians' values into health care provision.^{6 ,14} Nonetheless, the study by Piers et al³ raises the possibility that clinicians' values may already be influencing the provision of critical care, and that heterogeneity in such values may result in undue variations in practice.

Although the report by Piers et al³ provides a hazy lens through which to view appropriateness of care, it yields more clarity than prior studies. Thus, the greatest contribution of APPROPRICUS may be to provide the clarion call needed to spur more rigorous study of what happens to clinicians and the care they provide when requests for care do not resonate with clinicians' conceptions of appropriateness. Such clinician-centered outcomes research, in other words, may usefully supplement the patient's perspective in gauging the quality of health care delivery.