Introduction

Shared decision-making (SDM) is a collaborative process in which patients and providers partner to make informed decisions in line with patients’ needs and priorities [1]. This entails (1) informing patients that options are available, (2) providing more detailed information on the options, and (3) supporting patients in exploring their preferences and deciding what is best for them [1]. SDM is imperative in preserving patient autonomy and delivering high-quality cancer care [2, 3]. However, for patients newly diagnosed with advanced cancer, SDM is challenging due to the amount and complexity of information and undesirable emotions associated with the diagnosis [4, 5].

Prior studies have found that patients diagnosed with cancer and their caregivers have a high level of unmet information needs, particularly surrounding treatment benefits and side effects [4, 6]. However, there is a limited understanding about the information and decisional needs of patients who are diagnosed with advanced cancer in the hospital as they await establishing outpatient oncology care. Hospitalized patients may face greater information needs due to unique barriers related to hospitalization, such as acute illness, inpatient-outpatient transitions, and worse survival outcomes [7]. Understanding these needs and barriers would inform potential inpatient interventions to help patients engage in SDM in the clinic, thereby enhancing the overall quality of cancer care provided to this vulnerable population.

Here, we conducted a mixed-methods study to understand the information and decisional needs of these patients. We also explored barriers/facilitators to these needs, particularly those related to their hospitalization. We focused on patient information needs, as opposed to the deliberation of preferences with one’s physician that is crucial in SDM, as the latter cannot be observed in hospitalized patients who have yet to meet their outpatient oncologist.

Methods

Study design

A convergent-parallel mixed-methods study of patients newly diagnosed with an advanced solid malignancy in the hospital, their caregivers, and outpatient oncologists was conducted at the University of California, San Francisco (UCSF) from 6/2022–11/2022. A mixed-methods approach was chosen to obtain rich, contextualized insights and validate findings across multiple methods. Patients, caregivers, and oncologists were selected because they play important roles in patient decision-making based on the Ottawa Decision Support Framework (ODSF) [8, 9]. ODSF conceptualizes the support needed by patients, families, and providers for difficult decisions. For this study about patients newly diagnosed with advanced cancer in the hospital, this decision was the most important anticipated cancer-related decision identified by these patients [8, 9].

Patients were consecutively sampled in person among inpatients for whom medical oncology consultation was requested. Eligibility criteria included (1) English-speaking adults, (2) clinical suspicion for newly diagnosed unresectable/metastatic non-hematologic malignancy (biopsy-confirmed or in-process with preliminary results), (3) no inpatient systemic therapy, and (4) no plan for hospice care. Patients had been told of their cancer diagnosis by a member of either the primary or consulting oncology team prior to being approached by the study team. Patients had the option of inviting a caregiver, who was then approached by the study team to participate in dyadic interviews. All medical oncologists who provide ambulatory care for patients with non-hematologic cancers at UCSF were approached. Patients and oncologists were reimbursed with $20 and $10 gift cards, respectively; caregivers were not reimbursed. The COREQ checklist for reporting qualitative research was followed (Supplementary Table 1) [10].

Data collection

Patient characteristics

To characterize the patient sample, patients were administered a survey at discharge with questions about demographics, whether they know their cancer type and stage, and decisional conflict using the 16-item decisional conflict scale (Supplementary Fig. 1) [9].

Patients and caregivers: qualitative data collection

Semi-structured interviews with patients ± caregivers were conducted 7–14 days post-discharge. An interview guide was developed and pilot-tested using the Ottawa Hospital Research Institute’s Decisional Needs Assessment, a data collection strategy based on the ODSF to identify the needs of patients to make better decisions (Supplementary Fig. 2) [11]. Questions were about patients’ understanding of their cancer, their most important upcoming cancer-related decision, the information and decisional support needed to make this decision, and barriers/facilitators to these needs. Interviews were conducted by telephone or virtual conference, audio-recorded, and transcribed by two co-authors. The sample size for the initial analysis was set at 8, and participants were interviewed until saturation was met, defined as no new themes in two consecutive interviews [12].

Oncologists: quantitative and qualitative data collection

Oncologists were e-mailed a pilot-tested survey that instructed them to recall the last appointment they had with a patient who was newly diagnosed with advanced cancer in the hospital and then asked how well-informed the patient was about cancer type, stage, treatment, and prognosis (4 items, 5-point Likert scale). The survey also included three open-ended questions about (1) cancer-related information patients should know before their initial appointment, (2) recommended cancer-related resources, and (3) suggestions for inpatient oncologists to facilitate SDM for patients at their future outpatient oncology visit (Supplementary Fig. 3).

Data analysis

Quantitative data

Quantitative data were summarized using descriptive statistics. The proportion of oncologists who reported their patient was very/somewhat poorly informed was compared pairwise for each information domain using the Chi-squared test, with “cancer type” as the reference domain. Statistical analyses were conducted using R 4.2.1 software with p < 0.05 deemed statistically significant.

Qualitative data

Constant comparison thematic analysis was conducted using a deductive approach while remaining open to new themes. Analysis was focused on themes that were unique to or amplified by hospitalization. An initial codebook was created using the interview guide and updated during the coding process using constant comparison. Transcripts were first coded by K.R.R. or P.W. using ATLAS.ti; then, secondary coding was performed by the other co-author. Discrepancies were negotiated to a consensus. Unresolvable discrepancies were addressed by D.H.K. Codes were then categorized into themes and subthemes by K.R.R. and D.H.K.

Quantitative–qualitative data integration

Quantitative and qualitative results were reported separately. During interpretation, K.R.R. and D.H.K conducted methodological triangulation by comparing qualitative patient and caregiver interview data with quantitative and qualitative oncologist survey data to look for similarities and disagreements across findings from the different methods to gain a more complete understanding. Interpretations were described in the Discussion through a weaving narrative approach, in which quantitative and qualitative results were reported on a concept-by-concept basis [13].

Results

Patient and caregiver characteristics

Saturation was met after 12 interviews, which includes 12 patients and 4 caregivers. Eighteen patients were approached to yield this sample size (response rate 67%). Median age was 72.5 years. Seven (58%) identified as male and 5 (42%) female; 5 (42%) identified as Non-Hispanic White (Table 1). All caregivers were female; two were daughters (50%), one (25%) sister, and one (25%) wife. Nine (75%) patients reported knowing their cancer type, and five (42%) knew the stage. Eight (67%) patients had a decisional conflict scale score > 37.5 (scored 0–100), which correlates with decision delay or uncertainty [11].

Table 1 Patient characteristics
Full size table

Patient and caregiver: qualitative results

Table 2 illustrates the full thematic analysis. Themes are described below in narrative form below Table 2, with patient quotes indicated by “P” and caregiver quotes by “C.”

Table 2 Thematic analysis of semi-structured interviews with patients and their caregivers about information and decisional needs
Full size table

Anticipated cancer treatment decisions

Patients and caregivers reported that the most important anticipated cancer decision involved treatment: whether to receive treatment, treatment type, and how aggressively to treat cancer. Patients described experiencing decisional conflict due to insufficient information about the cancer, pending test results from hospitalization, and not understanding potential treatment effects.

[Anticipated decisions about the cancer] are what I gotta figure out. How can I know the future unless I know what the potential is?…I gotta know what it is before I can even think about it. (P2)

Variable control preferences in decision-making

Decision-making control preferences varied from self-, caregiver-, or provider-directed decisions. Most patients desired full autonomy, while some deferred decisions to their future oncologist. Two patients relied on family to make decisions due to impaired decision-making capacity from illness severity, including one patient recovering from emergent diverting colostomy surgery.

[Caregiver’s brother] and I have been making the decisions for her since she has been too sick to be able to. (C1)

High cancer-related information needs and uncertainty

After hospital discharge, patients and caregivers reported known and unknown information needs. Some expressed uncertainty and anxiety about their cancer care both during and after the hospitalization. One caregiver described their worries from not knowing the cancer type or treatment options during a prolonged hospitalization.

We stayed in the hospital for four or five days and nothing happened. Nobody told us what it was and nobody told us we were going to start treatment. We’re just sitting there worrying we’re going to die because we’re not going to get treatment. (C2)

Participants desired to learn about the nature of the cancer; treatment details about options, benefits, risks, and urgency; and symptom management. Prognostic information was also highly desired, such as anticipated disease course and curability. Sometimes, patients and caregivers struggled to understand why test results were delayed. One caregiver stated it would have been helpful to be informed of the reasons for the delay of biopsy results. Lastly, patients described a lack of understanding of the next steps in cancer care after discharge.

Barriers to information gathering during hospitalization

High information quantity and complexity, along with a lack of information due to pending or delayed test results, hindered patients’ ability to understand their cancer. The flow of information also posed a barrier. For instance, one patient who was transferred from another hospital was confused after receiving contradictory diagnoses from different hospitals about whether their cancer was small-cell or non-small-cell lung cancer. Patients also reported difficulty feeling informed due to the dynamic, rapidly collected nature of the information being gathered, synthesized, and presented to them in the hospital.

Everybody was still learning. … And they say we’ll just keep passing the information as it comes to us. … And each day was a little bit different. (P4)

There were also multiple inpatient providers delivering information, which made it difficult for patients to track and reconcile information from multiple sources.

Caregiver: When you had lots of different doctors visiting, did you know who all of them were and what they were asking you? Patient: Sort of. Some I might know; some I don't know. (C3 & P3)

Additionally, emotional distress from the cancer diagnosis and physical distress from cancer-related symptoms made it difficult for patients to absorb information in the hospital.

[In the hospital], I was just in a lot of pain and under a lot of medication and had difficulty learning some new skills as to how to navigate this new body. (P5)

Facilitators to information gathering during hospitalization

A number of factors helped patients gather cancer-related information in the hospital, including rapid communication of test results and availability of hospital providers to share information. Caregivers helped explain information, especially for patients whose illness compromised information processing.

Without our sister being able to explain all of the tests, I think it would have been a lot more difficult...Our sister is a doctor and is able to read her chart. She can relay that information, and that helped a lot. (C3)

The hospitalization itself, which expedited cancer diagnostic work-up, was also viewed as educational.

We understand sometimes it might be hard with the hospital since a lot of things are happening all at once. They’ve made it as easy as possible … [The hospitalization] was an education. (P11)

Barriers to information gathering post-hospitalization

Patients continued to face challenges in gathering and processing information after discharge but before seeing an outpatient oncologist. Some had difficulty integrating the tremendous amount of scattered data obtained during the hospitalization. Moreover, patients and caregivers faced delays in scheduling an appointment quickly with an outpatient oncologist, sometimes due to insurance coverage problems.

After getting discharged, the insurance was the confusing part to try to see how we could continue care at UCSF. (C3)

Facilitators to information gathering post-hospitalization

The primary means by which patients sought to obtain more information about cancer treatment after being discharged was by quickly meeting with an outpatient oncologist. Patients also sought information from other sources, such as the Internet, friends, and other physicians.

Decision-making facilitators

Multiple factors helped patients prepare to make cancer treatment decisions beyond the educational benefits of obtaining information. One patient said the hospitalization gave him a better sense of the cancer and future directions, which reduced his distress.

[The hospitalization] has taken some of the pressure and anxiety away. [I went] from not knowing to having some idea of where I am, where my body is, and what the possibilities and my future existence are. (P1)

For other patients, returning home to a more comfortable environment after being in the hospital helped them prepare for their anticipated cancer decisions.

Oncologists: quantitative results

Of 46 oncologists contacted, 32 (70%) with a median 6 years of clinical experience (range 1–40 years) completed the survey. When asked about their last patient with a newly diagnosed advanced cancer in the hospital seen in their clinic, 18 (56%) oncologists reported that patients were somewhat or very poorly informed about treatment options, greater than the 5 (16%) for cancer type (p < 0.001; Fig. 1). The proportion for cancer prognosis (n = 18, 56%) was also greater than cancer type (p < 0.001). There was no difference in the proportion of oncologists who reported patients were somewhat or very poorly informed about cancer type versus cancer stage (n = 11, 34%; p = 0.07).

Fig. 1
figure 1

Oncologists’ perspectives on cancer-related knowledge of their most recent patient with advanced cancer newly diagnosed in the hospital. This bar graph illustrates how well-informed oncologists viewed their most recent patient with advanced cancer newly diagnosed in the hospital. From left to right, the domains of cancer information are cancer type, cancer stage, treatment options, and prognosis. On pairwise comparison to cancer type, more oncologists reported that patients were very or somewhat poorly informed about treatment options and prognosis compared to cancer type, but not about cancer stage versus cancer type

Full size image

Oncologists: qualitative results

Open-ended survey responses, with oncologist quotations indicated below by “O,” revealed themes describing cancer-related information and setting expectations regarding patients’ cancer and care plan.

Variable expectations about patient knowledge after hospitalization

Prior to the initial oncology appointment, oncologists reported that patients should know their cancer type, stage, and treatment (general categories, types of systemic therapy, toxicities, and curative versus palliative intent). Some reported that prognosis, such as likelihood of cure and longevity, should be disclosed early to facilitate information consolidation and outpatient consultation. Techniques to disclose prognostic information were described, such as best case-worst case framing and giving “broad strokes” ranges of survival.

It is super helpful for patients to know high level “maybe curable” or “probably not curable” so that they are emotionally prepared to listen and ask the right questions. If they hear this for the first time at clinic visit, there is often a “deer in the headlights” effect that interferes with productive consultation. (O5)

Other oncologists did not expect patients to know much about the cancer.

I’m okay with my patients having a very blank slate that I can fill in. (O6).

One oncologist believed treatment or prognostic information in the hospital is not helpful because it is frequently incorrect.

I don’t think treatment or prognosis information is helpful as the information they receive in the hospital is frequently inaccurate for their specific cancer subtype. (O3)

Facilitators to patient decision-making

Many oncologists wrote that inpatient oncology education is crucial in promoting SDM in the clinic, as repetition helps patients absorb and consolidate important information.

Patients often need to hear this information multiple times—it is so overwhelming at first, particularly if they are hearing it while unwell in the hospital. If they can come away knowing their diagnosis, that it is metastatic and what this means, and that there are treatment options, I think this is helpful. (O11)

Oncologists also recommended that inpatient providers highlight key cancer information in patients’ discharge instructions, such as cancer type, stage, and prognosis, provide reputable sources of online information (and what is unreliable), explain the roles of academic and community cancer centers regarding clinical trials and treatment location; describe the different roles of the multidisciplinary cancer care team, advise on the nature of the initial outpatient oncology visit, and suggest they bring in a trusted family member for support to oncology appointments.

Would be ideal to have [cancer type, stage, and rough prognosis] written down for them in discharge paperwork because patients often forget when they see doctors early AM/late PM/without family around/right after procedural sedation/etc. (O5)

To help outpatient oncologists communicate with these patients, they recommended inpatient providers document what information was or was not discussed with patients in the hospital, or give a handoff that includes pertinent inpatient data, such as inpatient care received, what information was given to the patient, and pending test results. Lastly, systemic changes to facilitate timely establishment of outpatient care were mentioned.

Patient uncertainty before outpatient oncologist appointment

Oncologists also acknowledged the significant uncertainty patients face while waiting for their outpatient oncology appointment following hospital discharge.

Discussion and conclusion

Discussion

In this mixed-methods study of patients, caregivers, and oncologists, we found that patients with advanced cancer newly diagnosed in the hospital have numerous information and decisional needs, experience decisional conflict, and face barriers to fulfilling these needs, many of which are unique to or exacerbated by hospitalization.

The interviews and surveys revealed insights on these patients’ information and decisional needs prior to meeting their outpatient oncologist. We found that typically, patients had an idea of their cancer type and stage, but most were unclear about treatment and prognosis, which is supported by our oncologist survey results and findings reported in the outpatient setting [14]. This cascade of knowledge follows the order of operations of first diagnosing and staging the cancer and then determining treatment and prognosis. Patients’ knowledge gaps about treatment and prognosis manifested as decisional conflict and uncertainty, which oncologists empathized as challenging for patients to experience as they await meeting their future oncologist.

Known barriers to information gathering and processing that were exacerbated by hospitalization contributed to these knowledge gaps. Barriers included emotional and physical distress from patients’ illness and low recall of complex, voluminous information, consistent with findings in studies in similar contexts [15,16,17,18,19]. Notably, patients experienced uncertainty due to inconsistencies in diagnostic test results as well as suboptimal provider-patient communication due to both a lack of information as well as the relaying of complex information about their cancer, which has not been previously reported in this context to our knowledge. The way such data were presented in the hospital was also uniquely challenging, such as difficulty reconciling information from numerous and constantly changing providers and lack of integration of inpatient test results following discharge. Lastly, all stakeholders recognized that establishing timely outpatient care was important in promoting SDM surrounding cancer treatment.

Despite the challenges of hospitalization, all participants described that the hospitalization also presented an opportunity to address patients’ information and decisional needs and aforementioned barriers. Beyond the benefits of rapid data gathering and reporting, hospitalization also alleviated decisional uncertainty since information can be relayed quickly and questions answered immediately by inpatient providers. This adds to previous studies that found that the ability to exchange desired information facilitates SDM in hospitalized patients [18, 20,21,22]. Oncologists also described how hospitalization can promote information retention and consolidation since inpatient providers can start a chain of information repetition, preparing patients to participate in SDM [23, 24].

Potential remedies to overcome the aforementioned barriers were identified. Consistent with what is known for patients with advanced cancer in the outpatient setting [8], actively involving caregivers was helpful in not only obtaining and consolidating information but also surrogating decision-making if needed. Oncologists recommended handoffs and tailored, standardized information in discharge instructions, which could address difficulties regarding pending test follow-up and a lack of information integration at discharge. This complements and adds to a recent qualitative oncology inpatient SDM study which proposed that the electronic health record could be leveraged to better promote SDM [18]. Lastly, education on reputable online information and orientation to cancer care delivery were recommended by both patients and oncologists to facilitate effective SDM.

There was a general concordance between patients and oncologists on the type of information patients ideally should know following discharge. However, there was concern for inaccuracy in treatment and prognostic information patients received in the hospital. This concern may have arisen from difficulty in providing accurate treatment and prognostic information with incomplete information and that patients’ performance status while hospitalized may not reflect their true baseline [25]. Also, inpatient providers are usually general oncologists who may have insufficient knowledge to educate patients on rare malignancies. For most patients, we suggest that inpatient providers communicate treatment and prognostic information in broad strokes based on available information, with contingencies based on pending information. This is because repetition of serious news may help information retention and increase emotional preparedness during treatment SDM. Furthermore, earlier prognosis disclosure and awareness of incurability of their cancer are associated with more accurate prognosis awareness and patient-directed decisions like hospice enrollment, respectively [23, 24]. Attention should be paid to prognostic subgroups that may vary based on pending tests, such as molecular testing, and it should be conveyed that this prognostic uncertainty will be clarified with their outpatient oncologist.

The single-center design, few caregivers, and exclusion of non-English-speaking patients limit the generalizability of our findings. Our study also did not explore patient preferences, which are important in SDM. Nonetheless, the mixed-methods design across three key decision-making roles and high patient and oncologist response rates strengthen our findings.

Conclusion

Patients diagnosed with a new advanced cancer in the hospital have numerous SDM needs, mainly surrounding treatment and prognostic information, and face decisional conflict surrounding anticipated treatment decisions. Barriers to SDM can be exacerbated by hospitalization, since acute emotional and physical distress limit cancer information absorption, and information is often high in volume and complexity and delivered by multiple, constantly changing providers. Despite these barriers, patients may become better prepared for SDM regarding their cancer because they have a concentrated time for information-sharing with providers and to quickly learn about their cancer, alleviating uncertainty and anxiety. Further investigation is needed to test potential interventions, such as those identified by our stakeholders, to address barriers and augment facilitators. Doing so may enhance SDM and promote high-quality cancer care for this vulnerable population.