Summary and reflection of findings
The dominant theme in the study data was that concerns about the absence of revenue, or indeed, the likely loss of revenue, were a major barrier preventing the implementation of SDM. Many other factors were prominent as well, but the view that SDM might impair organizational or individual profit margins and reduce the income of some health professionals was widespread. While the study focused on organizational- and system-level characteristics that promote or hinder SDM in US cancer care, this major theme also touches the individual clinician-patient level, as individual clinicians’ actions related to revenue generation were also discussed. However, they should not be villainized as greedy professionals, but seen as responding to the system in which they work.
On the organizational level, having leadership support for SDM within cancer care delivery organizations and multidisciplinary teams were viewed as critical contributors to potential future implementation. On the health system level, views diverged on whether embedding tools into EHRs, making SDM a criterion for accreditation and certification, and enacting legislation could promote the implementation of SDM.
Strengths and limitations
We do not claim, given the sample, that this qualitative study provides strong generalizable data. However, the purposive maximum variation strategy does provide a range of stakeholder views about the place of SDM in US cancer care. It is an inherent feature of this sampling approach that the selection of participants with diverse characteristics relies on the research team’s judgment. To minimize this effect, we consulted with key experts in our networks to identify potential participants. Yet, the selection might have been biased. Nevertheless, the identified themes are significant because they emerged out of this heterogeneous sample. At the same time, the chosen approach does not allow an investigation of differences between subgroups of participants, which has to be considered as a further limitation of this study. Of the purposively invited sample of stakeholders, 83% agreed to participate in an interview. In order not to compromise the anonymity of the participants, we did not gather extensive information on demographic and professional characteristics. It can be seen as a further limitation. However, this also ensured the necessary confidentiality for gathering sensitive data about how organizational- and system-level factors influence cancer care in the USA.
Comparison to previous work
Most previous studies that have examined the implementation of SDM have not reported the dominant role of revenue generation [
11,
21]. The US healthcare system may be particularly sensitive to impact on revenue generation. The US system differs from most OECD countries, being predominantly based on a fee-for-service structure, having low levels of purely salaried physicians [
22], and where a high percentage of GDP is spent on healthcare [
23]. Cancer specialists in the USA earn much more than in other countries [
24], and US oncologists charge more for office-administered drugs than other specialists [
25]. However, the impact on revenue may also apply in other contexts, as found in a Dutch study that describes financial interests of HCPs as one of many barriers towards SDM implementation in multidisciplinary sciatica care [
26]. New evidence that payment structures influence SDM implementation comes from an example in the Netherlands, where better and more patient-centered care at lower costs was achieved by moving away from fee-for-service payment model, and by changing hospital structures and culture [
27].
Furthermore, the rising costs of cancer treatment, although a challenge in many countries, have been described as particularly problematic in the USA [
28,
29]. The combination of a comparatively high proportion of people without health coverage [
30] and rising out-of-pocket costs paid by insured patients with cancer [
31] lead to substantial financial toxicity in patients with cancer [
32]. This burden can be aggravated by the opacity of treatment costs, which was described in our study and which is uniquely relevant in the US health care system [
33]. Despite the challenges in identifying overall costs, out-of-pocket costs need to be addressed early on in treatment decision-making processes [
34], as they have been shown to influence patient preferences and treatment adherence [
35]. Two recent studies have shown that SDM tools can trigger cost discussions, but that more support is needed to help HCPs address treatment cost and financial distress adequately [
36,
37].
System-level changes, such as legislative mandates and accreditation requirements, were viewed with some skepticism, comparable to HCPs’ attitudes towards accreditation in general [
38]. At the same time, legal requirements for quality improvement strategies and adoption of accreditation programs are known drivers of change [
39,
40]. In fact, even in the USA, several healthcare and cancer-specific policies have recommended the implementation of SDM [
5,
41]. The US National Comprehensive Cancer Network (NCCN) support the use of SDM in their patient guidelines, e.g., [
42,
43], and the communication consensus guideline by the American Society of Clinical Oncology includes SDM [
44]. Nevertheless, these policies have had little impact, and no enforcement. In relation to reducing financial toxicity, it has been pointed out that health policy (e.g., price negotiations, value-aligned pricing strategies) and regulatory interventions (e.g., revision of drug approval regulations) could play a crucial role in eliminating low-value care, i.e., the use of expensive treatment options with minimal to no clinical benefits [
33]. Another suggested strategy highlights the role of cancer societies, which could develop cost-conscious clinical practice guidelines [
34]. Such changes could indirectly foster SDM, by mitigating the necessity of revenue generation as a barrier to SDM implementation.
A number of studies point to the important roles of organizational leadership and care coordination through multidisciplinary teams [
11], and our findings echo results of a study that found the quality of healthcare can be improved by supportive and visionary leadership [
45]. Several participants reported organizational leaders having only lip service commitment to SDM, illustrating the known phenomenon where patient-centered organizational statements represent rhetoric more than real commitment [
46].
Our results can also be compared to a multi-study analysis investigating attributes of context found relevant by HCPs in the implementation of evidence-based practices [
13]. The main theme of our study was also found as one of several core contextual factors in this secondary analysis of 145 interviews from 11 studies in different clinical contexts (not including oncology) [
13]. That study also revealed that regulatory and legislative standards were less commonly described attributes that varied in their relevance across clinical contexts [
13].
Implications for future research
Future SDM implementation research would benefit from comparing the impact of a multicomponent SDM implementation program located in different payment models. As recently pointed out by Roberts and colleagues [
47], the use of health economic evaluations in implementation science is limited. An economic evaluation of SDM implementation, possibly using a mixed methods design [
48], would shed light on this under researched area. Also, future SDM implementation efforts in US cancer care need to go further than interventions targeting only the clinician-patient level (e.g., training for individual HCPs). Our study generally highlights the need to further investigate outer setting variables in implementation studies. In order to do so, it is necessary to develop further tools to measure those variables [
13] and to distinguish between modifiable and unmodifiable variables [
49]. Last, but not least, future research on treatment decision making processes should further investigate the role of treatment costs and financial toxicity.