Provisional title. “Convergent pulmonary decline” is offered as a working label for a proposed framework, not as a diagnostic term.

Abstract

Background. Pulmonary complications are among the most frequent serious events in advanced cancer. The clinical literature treats them as discrete entities — infectious pneumonia, drug-induced pneumonitis, radiation lung injury, malignant pleural effusion, thromboembolism, cardiogenic edema, post-obstructive collapse — each with its own diagnostic criteria and management pathway. That organization is correct for diagnosis and treatment. Its consequences for explanation have received less attention.

Problem, stated as a hypothesis. Patients and the people around them appear frequently to end up without an assembled causal account: a set of accurate condition-specific statements that never resolves into an answer to “what happened and why.” This paper proposes that the difficulty is one of assembly and transmission rather than of missing knowledge or absent explanation. Condition-specific communication may be delivered competently at every point and still fail to yield an integrated account, because integration is nobody’s assigned task, because it must be reconstructed anew at each encounter, and because whatever integration does occur degrades as it passes outward from the bedside to family at a distance, friends, and community clergy.

Framework. Convergent pulmonary decline names a pattern in which heterogeneous pulmonary insults, occurring in a patient whose reserve is already reduced by malignancy, treatment, and comorbidity, act through a limited set of shared mediators — loss of functional lung volume, impaired gas exchange, inflammatory injury and capillary leak, failure of secretion clearance, and loss of cardiopulmonary reserve — which reinforce one another such that the trajectory exceeds what any single component predicts.

Approach. The paper states in advance the conditions under which its central claim should be judged false, proposes the scoping review that would test it, and treats its own framework as a hypothesis about communication requiring evaluation rather than as an established finding.

Cautions. The framework is not a syndrome, a diagnostic category, or a billing code. It does not assert that clinicians fail to explain, that respiratory deterioration in cancer is typically irreversible, or that multifactorial death implies error.

1. Introduction: The Assembly Problem

Two composite cases follow. Both are fictional constructions assembled to illustrate contrasting patterns; neither describes any individual patient, and identifying features have been altered or invented throughout.

Case A. An older woman with a long smoking history is treated for lung cancer. Over months she is told about the tumor, then about a pneumonia, then about fluid requiring drainage, then about a second infection. Each explanation is accurate. When she dies, the family’s account is short: she smoked for forty years and it caught up with her. The account is not wrong, but it has absorbed everything else — the treatment, the effusion, the sequence of infections — into a single moral fact about the patient. Relatives at a distance receive this version and nothing else.

Case B. A woman treated for years for metastatic breast cancer, with disease long controlled and current treatment directed at the liver, is admitted with breathlessness. The family’s attention has been on liver values for months; the lungs were never the front. She deteriorates over several days amid competing possibilities — infection, treatment-related inflammation, effusion, fluid overload — and dies. Family members and their pastor are left holding a set of statements about the liver, about the lungs, about an infection, and about treatment, with no account that connects them. Relatives, friends, and a congregation receive fragments at second and third hand.

Both families end without an integrated causal account, from opposite directions: in Case A a ready-made folk explanation crowded out the actual pathway; in Case B no explanatory frame existed at all and the one in use pointed elsewhere.

Four observations set the problem, none of which requires supposing that anyone communicated badly:

1. The clinical literature describes each component thoroughly. Drug-induced interstitial lung disease, radiation pneumonitis, immune-related pneumonitis, malignant effusion, neutropenic infection, and cancer-associated thromboembolism each support mature evidence bases and, in most cases, formal guidance (Skeoch et al., 2018; Naidoo et al., 2017; Hanania et al., 2019; Feller-Kopman et al., 2018; Freifeld et al., 2011; Khorana et al., 2007).
2. Differential diagnosis properly separates them, since treatment differs sharply across them.
3. Integration is unassigned. Condition-specific communication has an owner — the clinician managing that condition. The integrated account across conditions, specialties, admissions, and months has no designated owner, no documentation slot, and no billing code.
4. Whatever integration occurs must survive relay. The person in the room is rarely the only person who needs the account, and frequently is not the person who will carry it to everyone else (Adams et al., 2009).

The paper’s question is therefore not “why don’t clinicians explain?” but:

Under what conditions does an integrated causal account get assembled, and by whom, and how far outward does it travel intact?

2. Scope and Terminology

2.1 What the framework names

A pattern with five features:

• one or more pulmonary insults occur, from causes that may be unrelated to one another;
• malignancy, prior treatment, and comorbidity have already lowered compensatory reserve;
• the resulting problems become mutually reinforcing rather than merely additive;
• recovery from any single component becomes harder as others accumulate;
• respiratory failure, where it occurs, extends into systemic deterioration.

The weight of the proposal rests on the second and third features. That insults occur and that patients sometimes die of them is unremarkable; the claim that diminished reserve converts survivable events into unsurvivable ones, and that the components feed one another, is what the framework exists to convey.

2.2 What it does not mean

It is not a disease entity, diagnostic code, or reimbursement category; not a substitute for determining whether infection is present; not an assertion that treatment caused every complication observed; not a prediction that any given episode is terminal; not a reason to withhold treatment of a reversible problem; and not a fixed sequence with predictable stages.

2.3 Distinctions commonly lost in relay

Table 1. Clinical terms and plain-language equivalents

Definitions of ARDS and sepsis follow the Berlin Definition and the Sepsis-3 consensus (ARDS Definition Task Force, 2012; Singer et al., 2016).

The single phrase “fluid in the lungs” covers at least three of these conditions, with different treatments and different prognostic weight. A family told that fluid was drained, and that the patient deteriorated anyway, may reasonably conclude that the drainage failed — when the drained pleural fluid and the undrained alveolar edema were separate problems. Terminological collapse of this kind survives relay far better than the distinctions do, because the collapsed version is shorter and easier to repeat.

2.4 The communication network beyond the dyad

Medical communication is designed for a dyad or a bedside group: clinician and patient, sometimes with a companion or a designated proxy. The audience that actually forms beliefs about a death is a network. It includes the patient; the family member present; family members at a distance; adult children coordinating by telephone; friends; congregations and their clergy; and coworkers and neighbors who will nonetheless offer opinions, comparisons, and sometimes advice that is relayed back inward.

This paper distinguishes three positions:

• Position 1 — in the room. Hears the clinician directly, can ask questions, can request clarification. Best served by existing practice.
• Position 2 — one relay out. Family and close friends receiving an account from Position 1, typically compressed, under stress, and without the ability to check it.
• Position 3 — two or more relays out. Extended family, friends, congregation, community clergy. Receives fragments, forms a settled account early, and rarely revises it.

Existing patient-education material, decision aids, and question prompt lists are written almost exclusively for Position 1 — the position with the least need of them, since that person can simply ask. No artifact in current practice is designed for Positions 2 and 3, which is where a large share of durable confusion, and most enduring attribution of blame, collects.

This is testable. If an integrated account is being delivered and simply not transmitted, interventions should target relay. If it is not being assembled in the first place, interventions should target assembly. If it is assembled and transmitted but rejected because it does not answer the question the family is actually asking — often a question about fault — then neither intervention will help, and the remedy lies elsewhere.

3. The Literature: A Hypothesis and Its Disconfirming Conditions

It would be easy, and unjustified, to assert that no integrated account of pulmonary decline in advanced cancer exists. Palliative medicine, critical care, and geriatric oncology all work with concepts adjacent to this one: functional trajectories at the end of life (Lunney et al., 2003; Gill et al., 2010), frailty and diminished physiological reserve, multifactorial dying, and uncertainty as a feature of serious illness (Mishel, 1988). Patient-centered communication in cancer care has been the subject of a sustained research program (Epstein & Street, 2007). The honest starting position is that the framework proposed here may already exist in substance under other names, or may exist in the professional literature but not in the material that reaches families.

Accordingly, the central claim is stated as a hypothesis with its disconfirming conditions specified in advance, following the logic of preregistration (Nosek et al., 2018).

Hypothesis (H1). In the literature addressed to patients and families, pulmonary complications of advanced cancer are presented predominantly one at a time, and integrated accounts of their interaction are rare.

H1 should be judged false if a scoping review finds any of the following:

• a body of patient-facing material that routinely presents two or more pulmonary mechanisms in interaction, rather than sequentially or in isolation;
• established clinical guidance assigning responsibility for an integrated causal account to a named role, with a documentation locus;
• an existing named construct in palliative or critical-care literature that performs the same explanatory work, whatever its label, with demonstrated penetration into patient-facing use;
• evidence that families ordinarily do arrive at integrated accounts, in which case the object of study is the exception rather than the rule.

A partial disconfirmation is anticipated and would be informative. If integrated accounts prove common in the professional literature but rare in what reaches Positions 2 and 3, then the object of study is not an unnamed phenomenon but a transmission and measurement problem — a more tractable finding, and one that would redirect the paper toward relay rather than toward conceptual novelty.

Proposed method. A scoping review following established framework and reporting standards (Arksey & O’Malley, 2005; Tricco et al., 2018), sampling three distinct literatures — professional clinical guidance, patient-education material issued by cancer centers and charitable organizations, and palliative-care communication research — with each source coded for the number of mechanisms addressed, whether interaction is described, the assigned owner of integration if any, and the intended audience position (1, 2, or 3).

Until that review is done, everything that follows should be read as a proposed framework awaiting a test of whether it is needed.

4. Initiating Pathways

These are described separately because they demand different treatment. Section 5 describes what they share.

4.1 Infectious injury

Susceptibility arises through several independent routes: neutropenia following cytotoxic therapy, with risk rising with depth and duration (Freifeld et al., 2011); impaired cellular and humoral immunity from the malignancy itself; corticosteroid exposure, often prescribed for an earlier pulmonary problem; bacterial, viral, and fungal pneumonias, including organisms uncommon in immunocompetent hosts; aspiration from dysphagia, obstruction, sedation, delirium, or weakness (Mandell & Niederman, 2019); post-obstructive infection distal to an airway narrowing (Ernst et al., 2004); and device-associated infection.

Infection may be the first event or a late complication of something else. This dual position is central: infection is both an entrance to the pathway and an exit from it.

4.2 Direct treatment-related toxicity

“The chemotherapy damaged her lungs” may denote drug-associated pneumonitis; interstitial lung disease with fibrotic change; diffuse alveolar damage histologically indistinguishable from ARDS; pulmonary hemorrhage; capillary leak with non-cardiogenic edema; or cardiac injury producing cardiogenic edema months to years later (Limper, 2004; Cardinale et al., 2020). Their treatments and trajectories differ.

4.3 Radiation-induced lung injury

Injury occurs in two phases: early radiation pneumonitis, inflammatory and often corticosteroid-responsive, and later pulmonary fibrosis, which does not reverse. Severity relates to dose, irradiated volume, fractionation, concurrent systemic therapy, and preexisting lung disease (Marks et al., 2010; Hanania et al., 2019). The point relevant here is the residue: after acute inflammation resolves, the patient retains permanently less usable lung — a deficit invisible in ordinary life and decisive when a new insult arrives.

4.4 Immune-mediated injury

Checkpoint inhibition produces pneumonitis in a minority of patients, with incidence varying by agent, combination, and tumor type (Nishino et al., 2016; Naidoo et al., 2017). Imaging patterns overlap with infection and with progression; distinguishing them at presentation is often impossible, so both are commonly treated at once; corticosteroids and second-line immunosuppression are the mainstay for the pneumonitis; and that treatment raises infection risk.

The resulting paradox — the drug controlling the inflammatory injury opens the door to the infectious one — should be stated plainly to families, because it explains a sequence of improvement followed by deterioration that otherwise looks like an error.

4.5 Targeted agents and antibody–drug conjugates

“Targeted” describes selectivity, not exemption. Molecular targets are often expressed in healthy tissue; cytotoxic payloads can affect bystander cells; and several agents carry recognized interstitial lung disease risk requiring monitoring and early intervention (Tarantino et al., 2021; Kroschinsky et al., 2017).

4.6 Cancer-mediated structural injury

Pulmonary metastases occupying parenchyma; lymphangitic spread producing dyspnea out of proportion to visible disease; pleural involvement and malignant effusion (Feller-Kopman et al., 2018); central airway obstruction with distal collapse (Ernst et al., 2004); impaired lymphatic drainage, which both generates fluid and prevents its clearance; trapped lung, where a fibrous pleural peel prevents re-expansion after drainage (Huggins et al., 2007); and chest-wall or respiratory-muscle involvement.

4.7 Systemic and comorbid contributors

Heart failure; renal dysfunction; hypoalbuminemia and malnutrition; anemia, reducing oxygen delivery at any given saturation; pulmonary embolism (Khorana et al., 2007); preexisting COPD or interstitial lung disease; frailty, sarcopenia, and deconditioning (Gill et al., 2010); impaired swallow and weak cough. None of these is a lung disease. All of them determine how a lung disease ends.

4.8 The unattended front

A distinct and, in the author’s view, under-described contributor deserves its own heading.

When metastatic disease is being actively managed at one site, clinical attention, imaging, laboratory surveillance, and the family’s mental model all calibrate to that site. A patient with breast primary and hepatic metastases is followed with liver function tests and abdominal imaging; the conversation is about the liver; the family learns to read liver values. Meanwhile the lungs may be accumulating quiet deficits — a small effusion, prior radiation change, cardiac strain from earlier anthracycline exposure, deconditioning, a weakening cough — none of which crosses a threshold for attention on its own, and none of which is anyone’s assigned surveillance responsibility.

Two consequences follow. Clinically, the first recognized pulmonary event in such a patient may occur against a background of reserve loss that was never measured, so that an apparently modest insult produces a disproportionate deterioration. Communicatively, the family’s expectation is pointed at the wrong organ. They have been told, accurately, that the situation is managed and that the liver is the concern. When death arrives through the lungs, it arrives outside every frame they have been given.

This suggests a practical measure: for patients with metastatic disease at any site who have prior thoracic radiation, prior pulmonary toxicity risk, cardiac exposure, recurrent effusion, or declining function, a periodic explicit assessment of respiratory reserve — and a corresponding sentence in the family conversation naming the lungs as a potential front even while the treatment focus lies elsewhere.

5. The Common Physiological Mediators

The proposal is to regroup Section 4 by consequence rather than by cause. Five mediators account for most of the convergence.

5.1 Loss of functional lung volume

Lung ceases to participate in gas exchange when it is compressed (effusion), filled (consolidation, edema), closed (atelectasis), replaced (tumor), or scarred (fibrosis). The mechanisms are unrelated; the arithmetic is identical.

Different problems can take away the use of parts of the lung. The body does not distinguish which one did it. What matters is how much working lung remains.

5.2 Impaired gas exchange

Ventilation–perfusion mismatch, diffusion impairment, and true shunt all reduce oxygenation, and they respond differently to supplemental oxygen — shunt physiology notably poorly (Parshall et al., 2012).

Blood keeps flowing through parts of the lung that are no longer receiving fresh air, or where inflammation, fluid, or scar stands between the air and the blood.

5.3 Inflammation and vascular permeability

Infection, radiation, cytotoxic drugs, immune activation, and sepsis converge on increased pulmonary capillary permeability. At the severe end this is ARDS (ARDS Definition Task Force, 2012), which in patients with malignancy carries a poor prognosis (Azoulay et al., 2014). A consequence worth stating: once established, the inflammation may make its own origin undiscoverable, because the end state looks similar regardless of what began it.

5.4 Failure of pulmonary clearance

Weakened cough from sarcopenia, pain, pleural disease, or exhaustion; retained secretions; airway obstruction; sedation; dysphagia and recurrent aspiration; fatigue and recumbency. Clearance failure converts an isolated infection into a recurrent one, and a recurrent one into a permanent one.

5.5 Loss of cardiopulmonary reserve

This concept bridges clinical and lay understanding better than any other and deserves explicit teaching.

A healthy person has far more lung than daily life requires. Losing part of it may not be noticeable. A person with cancer who is anemic, has lost muscle, has had prior lung injury, whose heart is under strain, and who is receiving active treatment may have very little unused capacity left. For that person, the same loss is the difference between breathing and not breathing.

This one idea accounts for most of what families find inexplicable: why a “small” pneumonia proved fatal; why the patient seemed stable until abruptly she was not; why drainage helped for a day; and why a complication a neighbor survived was not survivable here.

5.6 Systemic consequences

Hypoxemia → increased work of breathing → respiratory muscle fatigue → cardiac strain → delirium and exhaustion → (with infection) sepsis (Singer et al., 2016) → ARDS → acute kidney injury and circulatory failure → multiorgan dysfunction. At this stage the question of pneumonia versus pneumonitis has become therapeutically marginal, though it remains important to the family and to the death certificate.

6. The Self-Reinforcing Cascade

Figure 2. The convergent cascade.

flowchart TD
    A["Cancer, treatment, infection, or comorbidity"] --> B["Initial lung insult"]
    B --> C["Reduced respiratory reserve"]
    C --> D["Poor oxygenation and impaired clearance"]
    D --> E["Infection, inflammation, fluid, or collapse"]
    E --> C
    E --> F["Respiratory and systemic failure"]

The loop from E back to C carries the argument. Representative feedback paths: inflammation increases capillary leak, worsening oxygenation, increasing inflammatory stress; obstruction prevents an infection from clearing, so antimicrobials suppress without curing; infection provokes inflammatory injury beyond the infected region; distress exhausts weakened muscles, reducing tidal volume and cough; corticosteroids control pneumonitis while raising susceptibility to the next infection; sepsis impairs cardiac and renal function, adding edema to an already flooded lung; hypoxemia and infection produce delirium, raising aspiration risk.

Two implications follow. Timing matters more than in a linear model: an intervention sufficient at one point may be insufficient two days later, not because it stopped working but because the system it entered has changed. And the search for a single cause becomes less meaningful as the number of active loops increases.

7. A Two-Axis Typology

Existing description sorts these cases by mechanism. The proposal here adds a second axis — what the family was primed to expect — because the two axes fail differently and require different remedies.

7.1 Axis I: mechanism-dominant patterns

Type I — Infection-dominant. Immunosuppression, neutropenia, or aspiration → pneumonia → sepsis or ARDS. Usually identifiable; often rapid.

Type II — Inflammation-dominant. Radiation, drug, or immune pneumonitis impairs oxygenation. Corticosteroid treatment succeeds, fails, or succeeds while permitting secondary infection. Survivors may retain fibrosis and a permanently lower ceiling.

Type III — Pleural-fluid dominant. Malignant or paramalignant effusion compresses lung. Drainage may give real relief or be limited by trapped lung or rapid reaccumulation (Feller-Kopman et al., 2018; Huggins et al., 2007).

Type IV — Obstruction-dominant. Tumor blocks an airway or lymphatic drainage, producing collapse and recurrent post-obstructive infection in the same location (Ernst et al., 2004).

Type V — Cardiovascular and fluid dominant. Cardiotoxicity, heart failure, renal dysfunction, hypoalbuminemia, or capillary leak produces edema and hypoxemia, often fluctuating (Cardinale et al., 2020).

Type VI — Mixed terminal decline. Several modest problems combine in a patient with minimal reserve. No component explains the outcome; the combination does.

Type VI most often generates the belief that something was missed. Where no single problem was large, families reasonably ask why no single problem was fixed. The answer is that the deterioration arose from interaction, and that this is a recognized pattern rather than an oversight.

7.2 Axis II: expectation patterns

Pattern A — Pre-supplied folk cause. A ready-made causal story exists before any pulmonary event — most commonly a smoking history with lung cancer, but also obesity, occupational exposure, or delayed presentation. The story is not false, and it answers the causal question so economically that demand for any further account collapses. The medical pathway is absorbed into a single fact about the patient’s history, and, frequently, into retrospective moral judgment of the patient (Chapple et al., 2004). Position 3 receives only this version.

Failure mode: explanation is crowded out rather than absent. Additional information is not sought and, if offered, may not be retained, because the question feels closed.

Pattern B — Displaced expectation. Attention has been calibrated to a different site or a different problem (Section 4.8). The family’s model is accurate but pointed elsewhere. Death arrives through an organ that was never the subject.

Failure mode: no frame exists, and the frame in use actively misleads. Families often report that they were told things were going well, which was true of the front being watched.

Pattern C — Diffuse or absent narrative. Long, complicated illness with many events, no dominant story, and no single clinician who has held the whole account. Families describe a blur.

Failure mode: assembly was never performed by anyone, and there is no one to ask for it now.

Pattern A needs displacement of an existing account; Pattern B needs anticipatory widening of the frame; Pattern C needs assembly and a designated owner. A single intervention will not serve all three, which is a reason to test them separately.

8. Diagnostic Uncertainty and Levels of Attribution

Competent care does not guarantee etiologic certainty. Symptoms are shared across mechanisms; imaging patterns overlap and rarely distinguish infectious from inflammatory infiltrate confidently; cultures have limited sensitivity; antimicrobials are appropriately started before sampling in a deteriorating patient; biopsy is often unsafe in hypoxemic or thrombocytopenic patients; infectious and non-infectious processes commonly coexist, so a positive culture does not exclude pneumonitis; not all fluid can be drained; and deterioration may outrun the diagnostic process.

The framework proposes a three-level attribution scheme, following the logic of death certification (Brooks & Reed, 2015):

1. Underlying cause — advanced malignancy and its cumulative burden, including the physiological cost of treatment.
2. Intermediate mechanisms — treatment-related lung injury, immunosuppression, obstruction, pleural disease, cardiac or renal dysfunction, malnutrition, frailty.
3. Proximate cause — the terminal event: pneumonia, ARDS, sepsis, respiratory failure.

“She died of cancer,” “the treatment injured her lungs,” and “pneumonia caused her death” are not competing claims. They are three levels of one causal history. Saying so explicitly, in those words, resolves a considerable share of post-death confusion at negligible cost — and it is short enough to survive two relays, which most explanations are not.

9. Treatment: Keeping Diagnostic Distinctions Inside the Unified Model

Convergence makes etiology harder to establish; it does not make it less consequential. Infection requires prompt, appropriately broad antimicrobial therapy, narrowed as data allow (Freifeld et al., 2011; Metlay et al., 2019). Immune-mediated pneumonitis requires corticosteroids and sometimes further immunosuppression, with attention to the infection risk created (Naidoo et al., 2017). Pleural effusion may require thoracentesis, indwelling catheter, or pleurodesis, chosen partly on whether the lung will re-expand (Feller-Kopman et al., 2018; Bibby et al., 2018). Edema may respond to diuresis and treatment of its cardiac or renal cause (Cardinale et al., 2020). Thromboembolism requires anticoagulation weighed against bleeding risk. Obstruction may be relieved bronchoscopically, by stenting, by radiation, or by tumor-directed therapy (Ernst et al., 2004). Dyspnea requires treatment as a symptom whatever its cause (Booth et al., 2008; Parshall et al., 2012).

Families should also be told why several of these are often given at once. Simultaneous empirical treatment of competing hypotheses is not indecision; it is the correct response to a patient deteriorating faster than the diagnostic sequence can proceed. Stated plainly, it reads as competence. Left unstated, it reads as guesswork — and the second reading is the one that travels outward.

10. Prognosis and the Meaning of Reversibility

Two questions are habitually conflated and should be separated explicitly:

1. Is the immediate complication treatable?
2. Does this patient retain enough reserve to recover if it is treated?

The first is frequently answered yes and the second frequently no. Conflating them produces both false hope and premature despair.

Bearing on the second: status and treatment-responsiveness of the malignancy; baseline performance status before the acute event (Prigerson et al., 2015); oxygen requirement and its trajectory; non-pulmonary organ dysfunction; recent admission frequency; recurrent effusion, particularly without re-expansion; inability to clear infection or taper corticosteroids; nutritional and functional decline preceding admission; and whether prior interventions produced recovery or only temporary stabilization.

Outcomes for patients with solid tumors admitted to intensive care have improved and should not be presumed uniformly poor (Puxty et al., 2014); ARDS in this population remains highly lethal (Azoulay et al., 2014). Both belong in an honest conversation, as does the finding that prognostic disclosure has not been shown to damage the clinical relationship or worsen distress in the way clinicians often anticipate (Enzinger et al., 2015).

A patient-facing formulation:

“We may be able to treat part of what is happening. Whether that helps depends on something separate — whether her lungs and the rest of her body still have enough reserve to respond. We should know more within a day or two, and I will tell you what I see.”

11. Assembly and Transmission

This section is the practical center of the revised framework.

11.1 Name an owner for the integrated account

Condition-specific explanation has an owner by default. The integrated account does not. The proposal is that one clinician — most naturally the treating oncologist or a palliative care clinician where involved — be identified as responsible for holding and periodically restating the whole picture, and that this account have a documentation locus in the record so that it survives handovers, admissions, and specialty transitions.

This is an organizational change, not a clinical one, and it costs little. Its absence is not anyone’s fault; it is a gap in role design.

11.2 Anticipatory triggers, widened

Existing anticipatory guidance is triggered by thoracic events. Adding triggers for the unattended front (Section 4.8):

• initiation of thoracic radiation;
• prescription of an agent with recognized pneumonitis or interstitial lung disease risk;
• a first or recurrent pleural effusion;
• identification of pulmonary metastases, lymphangitic spread, or airway involvement;
• a second pulmonary admission within a short interval;
• any new oxygen requirement or measurable decline in exercise tolerance, whatever the site of active disease;
• metastatic disease at a non-pulmonary site in a patient with prior thoracic radiation, prior cardiac exposure, or preexisting lung disease.

Early discussion of what may happen is associated with care better matched to patient preference and with improved bereavement outcomes, without the increase in distress often feared (Wright et al., 2008; Mack et al., 2012; Teno et al., 2004). Early palliative care involvement in advanced lung cancer has been associated with improved quality of life and mood (Temel et al., 2010; Kelley & Morrison, 2015).

11.3 Layered communication

Three layers, allowing the listener to stop at any point:

• Headline: “Several things are affecting her lungs at once.”
• Mechanism: “Inflammation, infection, and fluid are each reducing the amount of lung available to take up oxygen, and they make one another worse.”
• Prognostic meaning: “Because she has so little reserve left, treating one of them may not be enough to restore her breathing.”

Confirm understanding by teach-back rather than by asking whether there are questions (Schillinger et al., 2003).

11.4 Design for relay

Layered communication assumes a listener who will retain and reproduce it. Position 2 and Position 3 listeners will not, unaided. Practical measures:

• Give the headline in writing. Three or four sentences the family member can forward verbatim, dated, using the three attribution levels of Section 8. A written sentence survives relay; a spoken paragraph does not.
• Ask who else needs to know, and whether the person in the room is the one who will tell them. Frequently she is not, and frequently she is the least able to, being the most exhausted.
• Offer a second conversation for the wider circle — by telephone or video, at a scheduled time, where permitted by the patient. Family meetings have established practice guidance (Hudson et al., 2008); extending them outward is a modest step.
• Anticipate the fault question. If nobody supplies a causal account, one will be constructed, and constructed accounts tend to assign blame — to the treatment team, to a delay, or to the patient herself. Saying “this pattern is recognized, and it does not mean something was missed” is not defensive when it is true, and it is precisely the sentence that fails to reach Position 3 unless someone sends it there.

11.5 Community clergy and other non-institutional interpreters

Chaplaincy is inside the institution and appears in existing guidance (Puchalski et al., 2009). A congregation’s pastor is outside it. He receives everything at second or third hand, has no chart access and no standing to ask a clinician anything, and is nonetheless asked to interpret the death — at the bedside, at the funeral, and for a congregation, often with more durable effect on the family’s settled understanding than anything said in the hospital. Support from religious communities is common in advanced cancer and is associated with measurable differences in end-of-life care (Balboni et al., 2007, 2013), yet no part of the medical communication apparatus is designed to inform the community clergy who provide it.

Two low-cost measures, both requiring the patient’s permission: including the community clergy in the offer of a wider-circle conversation, and providing the same written headline given to the family. Neither requires disclosure beyond what the patient authorizes, and both address a position that currently receives only rumor.

11.6 Questions to encourage

For Position 1, at the bedside:

• Is the fluid around the lungs or inside them?
• Is this thought to be infection, treatment-related inflammation, cancer progression, or a combination?
• What evidence supports each possibility, and what would change your mind?
• Which parts of this are reversible?
• By when should we expect improvement if the treatment is working?
• How much respiratory reserve is left?
• What would tell you that recovery has become unlikely?
• Should palliative care be involved now?

For Positions 2 and 3, which the current literature does not address:

• Am I hearing about one problem or several?
• Is there anyone who has the whole picture, and can I hear it from them?
• Is what I was told about the liver (or the tumor, or the scan) still the main issue, or has something else become the main issue?
• Is there anything in writing I can read rather than reconstruct?

12. Ethical and Emotional Implications

No blanket verdict is warranted. Nothing in this paper establishes that clinicians fail to explain. The failure it describes — of assembly and transmission — is compatible with every individual communication having been accurate, appropriate, and well delivered. Structural gaps do not require culpable actors, and a framework of this kind should not be used to supply retrospective grievance. Where care was genuinely deficient, that is a separate question requiring separate evidence, and this framework neither supports nor forecloses it.

Honesty under uncertainty. Clinicians sometimes withhold an account because the cause is uncertain. The framework offers something to say when certainty is absent: the pattern can be described even when the trigger cannot be named. Silence is not neutral; families fill it, and what they fill it with is usually worse.

Blame directed at the patient. Pattern A cases carry a specific ethical hazard rarely discussed in this literature. Where a folk cause is available — most often smoking — the family’s account may resolve into moral judgment of the deceased, and patients themselves frequently anticipate and internalize this judgment (Chapple et al., 2004). Stating the actual pathway is not only explanatory but restorative: it returns the death to the domain of physiology from the domain of desert. Clinicians and clergy are both positioned to do this and neither is currently prompted to.

Hope and reserve are compatible. Hope for a reversible component and recognition of diminishing reserve are not alternatives, and presenting them as such forces a false choice.

Differing appetites for information. Some patients want detail, some very little, some want the family informed and themselves protected. The layered structure accommodates this without requiring anyone to guess in advance.

Multifactorial death is not evidence of error:

Recognizing a convergent terminal pathway does not mean treatment was futile or wrongly chosen. It means the benefits and burdens of treatment played out within a body whose capacity to absorb further injury had become very small.

13. Implications for Practice

1. A named owner of the integrated account, with a documentation locus (11.1).
2. A written three-to-four-sentence headline given to the family, designed for forwarding (11.4).
3. Standardized institutional plain-language definitions of pneumonia, pneumonitis, edema, and effusion, so that services do not use the same words differently.
4. A record prompt when two or more pulmonary processes are documented concurrently, suggesting a multifactorial-decline note and a family meeting.
5. Periodic explicit assessment of respiratory reserve in patients with non-pulmonary metastatic disease carrying pulmonary risk factors (4.8).
6. An offer of a wider-circle conversation, including community clergy where the patient wishes (11.5).
7. Explicit documentation of underlying, intermediate, and proximate causes, in notes and on the death certificate (Brooks & Reed, 2015).
8. A one-page diagram for families, with the Table 1 renderings alongside.
9. Discharge instructions naming warning symptoms, stating honestly what remains uncertain, and identifying whom to call.

Each is low-cost. The barrier is organizational, not scientific.

14. Research Agenda

Priority 1 — the disconfirming sweep. The scoping review specified in Section 3, conducted before further development of the framework. If H1 fails, the framework should be withdrawn or redirected.

Priority 2 — locating the failure. Where an integrated account does not reach a family, at which stage was it lost? A study sampling at four points — what was documented, what the clinician reports saying, what the Position 1 family member recalls at 48 hours and two weeks, and what a Position 3 relative or community clergy member reports understanding — would distinguish assembly failure from transmission failure. Degradation of narrative content through successive retellings is a well-established phenomenon in its own right (Bartlett, 1932); the question is its magnitude and shape here.

Descriptive questions. What proportion of respiratory admissions in advanced cancer involve two or more concurrent mechanisms? How often is a single etiology established with confidence before death? How accurately do death certificates represent multifactorial decline? How often does a pulmonary terminal event occur in patients whose active disease front was elsewhere?

Comprehension questions. Can families reliably distinguish pneumonia, pneumonitis, edema, and effusion after standard counseling? Does a one-page diagram improve comprehension over standard counseling? Does a written forwardable headline improve Position 3 understanding?

Outcome questions. Does the framework improve prognostic awareness without increasing anxiety or decisional conflict? Does early explanation affect intensive care utilization, hospice timing, or bereavement outcomes? Can features be identified marking the transition from reversible complication to convergent decline?

Typology questions. Do Patterns A, B, and C occur at rates justifying separate intervention? Does Pattern A predict patient-directed blame in bereaved family accounts?

Generalizability. Does the framework perform comparably across tumor types, treatment settings, languages, health literacy levels, faith communities, and cultural expectations about disclosure?

A first study. A two-arm pilot randomizing patients beginning a treatment with recognized pulmonary risk to standard counseling versus counseling supplemented by the diagram, the layered script, and the written forwardable headline — with outcomes of comprehension, decisional conflict, anxiety, and recall at two and eight weeks, and, where families consent, comprehension at one relay out.

15. Limitations

• Convergent pulmonary decline is a communication construct. It has not been validated and should not be treated as a clinical entity.
• The central claim about the literature is untested. Section 3 states the conditions under which it should be judged false; those conditions have not yet been evaluated.
• The relay model of Section 2.4, and the expectation typology of Section 7.2, are hypotheses generated from limited observation. They are offered as objects for study, not as findings.
• Cancer populations and treatments are highly heterogeneous, and the relative weight of the mechanisms described differs across tumor types and treatment eras.
• Pulmonary deterioration in advanced cancer is frequently reversible. The framework describes a possible trajectory, not a necessary one.
• Emphasizing convergence risks encouraging premature fatalism in clinicians as well as families, and must be paired with explicit attention to reversibility.
• Temporal sequence does not establish causation. That lung injury followed treatment does not show that treatment caused it.
• The author’s position is external to clinical medicine, with the specific consequences described in the author’s note below.

16. Conclusion

1. The initiating causes of pulmonary deterioration in advanced cancer are heterogeneous, clinically distinct, and separately consequential for treatment.
2. They converge through a small number of shared physiological mediators and mutually reinforcing complications, so that the trajectory may exceed what any single cause predicts.
3. An integrated causal account of that convergence is presently nobody’s assigned responsibility, and where it is assembled, no mechanism carries it beyond the person in the room.

The clinical task is to distinguish causes finely enough to treat whatever remains reversible. The communicative task is to assemble those causes into one account, and to send it far enough that the people who will carry the memory of this death actually receive it. The first task is well organized. The second, this paper proposes and does not yet demonstrate, is not organized at all.

Suggested Supplementary Materials

• Figure 1. Convergence diagram: eight initiating pathways converging on reduced respiratory reserve, branching to recovery or to progressive failure.
• Figure 2. The feedback cascade (Section 6).
• Figure 3. The relay diagram: Positions 1, 2, and 3, with existing communication artifacts mapped to show which positions they serve.
• Table 1. Clinical terms and plain-language equivalents.
• Table 2. The two-axis typology: mechanism patterns against expectation patterns, with the failure mode and indicated remedy for each cell.
• Box 1. The clinician script, the written forwardable headline template, and both question lists.

Author’s Note

Standing. I am not a physician, a nurse, or a researcher in any clinical field. I have no medical training, no patient contact, and no capacity to evaluate the primary evidence in this paper except as an attentive general reader. Everything in Sections 4 through 6 and 8 through 10 is a synthesis of secondary sources, and any reader with clinical training should assume I have made errors of emphasis or nuance in it and should check it accordingly. I have kept incidence figures out of the text for that reason, and I would rather this framework be corrected by people who know the material than published in a form that flatters my reading of it.

How this paper came to be written. It was prompted by two deaths from pulmonary complications in advanced cancer among people I knew — a relative at some distance, and a friend — in circumstances that differed in nearly every clinical particular. In neither case was I a person to whom clinicians spoke. In both cases I was two or three relays out, receiving accounts from family members and, in one instance, from a pastor, and in both cases what reached me was a set of accurate-sounding statements about particular conditions and treatments with no through-line connecting them.

What that vantage can and cannot support. It cannot support any claim about what clinicians said. I do not know. The residue I observed is consistent with an integrated account never having been offered, having been offered and not absorbed under stress, having been offered piecemeal by different people with no one assembling it, having been absorbed but not survived relay outward, or having been offered and rejected because it did not answer the question the family was actually asking. Those five possibilities are indistinguishable from where I stood and have almost nothing in common as remedies. This paper accordingly proposes a research program rather than a verdict, and Section 3 states in advance what would show its central claim to be wrong.

What the vantage does support is narrower and, I think, still worth something. If the concern is whether families arrive at an understanding of what happened, then the outer edge of the relay is the correct place to measure, and it is a place almost nothing samples. That observation does not require clinical training, and it is the one contribution here I am reasonably confident is mine to make.

On motive. Grief supplies conviction cheaply, and the conviction it supplies most readily is that something was withheld. I have tried to write against that. The clinicians in both cases may well have done everything correctly; nothing I saw is evidence otherwise, and I have removed from the paper every formulation that implied otherwise. If this framework is taken up, the use I would least want made of it is retrospective grievance against people who were treating a dying patient competently under uncertainty.

On the two cases. They function here as the prompt, not the evidence. They are described in the text only as fictionalized composites, with identifying features altered. Two observations can indicate where to look. They cannot establish that anything is missing, and the reader should not treat them as though they could.

Conflicts of interest and funding. None. The author has no financial relationship with any pharmaceutical manufacturer, health system, or professional body, and no clinical, academic, or commercial stake in the framework’s adoption.

Note on Sources

The references identify the major bodies of work supporting each section. Several are formal clinical guidelines, and guideline documents are periodically revised: the guidance on malignant pleural effusion, febrile neutropenia, community-acquired pneumonia, and immune-related adverse events should each be checked against the current version before this paper is submitted or used to inform practice. Incidence figures for pneumonitis associated with checkpoint inhibitors and antibody–drug conjugates continue to be refined as newer agents and combinations enter use, which is a further reason no such figures are stated in the text. Readers should verify all citations against the published record; a paper of this kind earns its standing by being checkable.

References

Adams, E., Boulton, M., & Watson, E. (2009). The information needs of partners and family members of cancer patients: A systematic review. Patient Education and Counseling, 77(2), 179–186. https://doi.org/10.1016/j.pec.2009.03.027

ARDS Definition Task Force. (2012). Acute respiratory distress syndrome: The Berlin definition. JAMA, 307(23), 2526–2533. https://doi.org/10.1001/jama.2012.5669

Arksey, H., & O’Malley, L. (2005). Scoping studies: Towards a methodological framework. International Journal of Social Research Methodology, 8(1), 19–32. https://doi.org/10.1080/1364557032000119616

Azoulay, E., Lemiale, V., Mokart, D., Pène, F., Kouatchet, A., Perez, P., Vincent, F., Mayaux, J., Benoit, D., Bruneel, F., Meert, A.-P., Nyunga, M., Rabbat, A., & Darmon, M. (2014). Acute respiratory distress syndrome in patients with malignancies. Intensive Care Medicine, 40(8), 1106–1114. https://doi.org/10.1007/s00134-014-3354-0

Back, A. L., Arnold, R. M., Baile, W. F., Tulsky, J. A., & Fryer-Edwards, K. (2005). Approaching difficult communication tasks in oncology. CA: A Cancer Journal for Clinicians, 55(3), 164–177. https://doi.org/10.3322/canjclin.55.3.164

Baile, W. F., Buckman, R., Lenzi, R., Glober, G., Beale, E. A., & Kudelka, A. P. (2000). SPIKES—A six-step protocol for delivering bad news: Application to the patient with cancer. The Oncologist, 5(4), 302–311. https://doi.org/10.1634/theoncologist.5-4-302

Balboni, T. A., Balboni, M., Enzinger, A. C., Gallivan, K., Paulk, M. E., Wright, A., Steinhauser, K., VanderWeele, T. J., & Prigerson, H. G. (2013). Provision of spiritual support to patients with advanced cancer by religious communities and associations with medical care at the end of life. JAMA Internal Medicine, 173(12), 1109–1117. https://doi.org/10.1001/jamainternmed.2013.903

Balboni, T. A., Vanderwerker, L. C., Block, S. D., Paulk, M. E., Lathan, C. S., Peteet, J. R., & Prigerson, H. G. (2007). Religiousness and spiritual support among advanced cancer patients and associations with end-of-life treatment preferences and quality of life. Journal of Clinical Oncology, 25(5), 555–560. https://doi.org/10.1200/JCO.2006.07.9046

Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge University Press.

Bibby, A. C., Dorn, P., Psallidas, I., Porcel, J. M., Janssen, J., Froudarakis, M., Subotic, D., Astoul, P., Licht, P., Schmid, R., Scherpereel, A., Rahman, N. M., Cardillo, G., & Maskell, N. A. (2018). ERS/EACTS statement on the management of malignant pleural effusions. European Respiratory Journal, 52(1), 1800349. https://doi.org/10.1183/13993003.00349-2018

Booth, S., Moosavi, S. H., & Higginson, I. J. (2008). The etiology and management of intractable breathlessness in patients with advanced cancer: A systematic review of pharmacological therapy. Nature Clinical Practice Oncology, 5(2), 90–100. https://doi.org/10.1038/ncponc1034

Boyd, K., & Murray, S. A. (2010). Recognising and managing key transitions in end of life care. BMJ, 341, c4863. https://doi.org/10.1136/bmj.c4863

Brooks, E. G., & Reed, K. D. (2015). Principles and pitfalls: A guide to death certification. Clinical Medicine & Research, 13(2), 74–82. https://doi.org/10.3121/cmr.2015.1276

Cardinale, D., Iacopo, F., & Cipolla, C. M. (2020). Cardiotoxicity of anthracyclines. Frontiers in Cardiovascular Medicine, 7, 26. https://doi.org/10.3389/fcvm.2020.00026

Chapple, A., Ziebland, S., & McPherson, A. (2004). Stigma, shame, and blame experienced by patients with lung cancer: Qualitative study. BMJ, 328(7454), 1470. https://doi.org/10.1136/bmj.38111.639734.7C

Elwyn, G., Frosch, D., Thomson, R., Joseph-Williams, N., Lloyd, A., Kinnersley, P., Cording, E., Tomson, D., Dodd, C., Rollnick, S., Edwards, A., & Barry, M. (2012). Shared decision making: A model for clinical practice. Journal of General Internal Medicine, 27(10), 1361–1367. https://doi.org/10.1007/s11606-012-2077-6

Enzinger, A. C., Zhang, B., Schrag, D., & Prigerson, H. G. (2015). Outcomes of prognostic disclosure: Associations with prognostic understanding, distress, and relationship with physician among patients with advanced cancer. Journal of Clinical Oncology, 33(32), 3809–3816. https://doi.org/10.1200/JCO.2015.61.9239

Epstein, R. M., & Street, R. L., Jr. (2007). Patient-centered communication in cancer care: Promoting healing and reducing suffering (NIH Publication No. 07-6225). National Cancer Institute.

Ernst, A., Feller-Kopman, D., Becker, H. D., & Mehta, A. C. (2004). Central airway obstruction. American Journal of Respiratory and Critical Care Medicine, 169(12), 1278–1297. https://doi.org/10.1164/rccm.200210-1181SO

Feller-Kopman, D. J., Reddy, C. B., DeCamp, M. M., Diekemper, R. L., Gould, M. K., Henry, T., Iyer, N. P., Lee, Y. C. G., Lewis, S. Z., Maskell, N. A., Rahman, N. M., Sterman, D. H., Wahidi, M. M., & Balekian, A. A. (2018). Management of malignant pleural effusions: An official ATS/STS/STR clinical practice guideline. American Journal of Respiratory and Critical Care Medicine, 198(7), 839–849. https://doi.org/10.1164/rccm.201807-1415ST

Freifeld, A. G., Bow, E. J., Sepkowitz, K. A., Boeckh, M. J., Ito, J. I., Mullen, C. A., Raad, I. I., Rolston, K. V., Young, J.-A. H., & Wingard, J. R. (2011). Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clinical Infectious Diseases, 52(4), e56–e93. https://doi.org/10.1093/cid/cir073

Gill, T. M., Gahbauer, E. A., Han, L., & Allore, H. G. (2010). Trajectories of disability in the last year of life. New England Journal of Medicine, 362(13), 1173–1180. https://doi.org/10.1056/NEJMoa0909087

Hanania, A. N., Mainwaring, W., Ghebre, Y. T., Hanania, N. A., & Ludwig, M. (2019). Radiation-induced lung injury: Assessment and management. Chest, 156(1), 150–162. https://doi.org/10.1016/j.chest.2019.03.033

Hudson, P., Quinn, K., O’Hanlon, B., & Aranda, S. (2008). Family meetings in palliative care: Multidisciplinary clinical practice guidelines. BMC Palliative Care, 7, 12. https://doi.org/10.1186/1472-684X-7-12

Huggins, J. T., Sahn, S. A., Heidecker, J., Ravenel, J. G., & Doelken, P. (2007). Characteristics of trapped lung: Pleural fluid analysis, manometry, and air-contrast chest CT. Chest, 131(1), 206–213. https://doi.org/10.1378/chest.06-0430

Kelley, A. S., & Morrison, R. S. (2015). Palliative care for the seriously ill. New England Journal of Medicine, 373(8), 747–755. https://doi.org/10.1056/NEJMra1404684

Khorana, A. A., Francis, C. W., Culakova, E., Kuderer, N. M., & Lyman, G. H. (2007). Thromboembolism is a leading cause of death in cancer patients receiving outpatient chemotherapy. Journal of Thrombosis and Haemostasis, 5(3), 632–634. https://doi.org/10.1111/j.1538-7836.2007.02374.x

Kroschinsky, F., Stölzel, F., von Bonin, S., Beutel, G., Kochanek, M., Kiehl, M., & Schellongowski, P. (2017). New drugs, new toxicities: Severe side effects of modern targeted and immunotherapy of cancer and their management. Critical Care, 21(1), 89. https://doi.org/10.1186/s13054-017-1678-1

Limper, A. H. (2004). Chemotherapy-induced lung disease. Clinics in Chest Medicine, 25(1), 53–64. https://doi.org/10.1016/S0272-5231(03)00123-4

Lunney, J. R., Lynn, J., Foley, D. J., Lipson, S., & Guralnik, J. M. (2003). Patterns of functional decline at the end of life. JAMA, 289(18), 2387–2392. https://doi.org/10.1001/jama.289.18.2387

Mack, J. W., Cronin, A., Taback, N., Huskamp, H. A., Keating, N. L., Malin, J. L., Earle, C. C., & Weeks, J. C. (2012). End-of-life discussions among patients with advanced cancer: A cohort study. Annals of Internal Medicine, 156(3), 204–210. https://doi.org/10.7326/0003-4819-156-3-201202070-00008

Mandell, L. A., & Niederman, M. S. (2019). Aspiration pneumonia. New England Journal of Medicine, 380(7), 651–663. https://doi.org/10.1056/NEJMra1714562

Marks, L. B., Bentzen, S. M., Deasy, J. O., Kong, F.-M., Bradley, J. D., Vogelius, I. S., El Naqa, I., Hubbs, J. L., Lebesque, J. V., Timmerman, R. D., Martel, M. K., & Jackson, A. (2010). Radiation dose–volume effects in the lung. International Journal of Radiation Oncology, Biology, Physics, 76(3, Suppl.), S70–S76. https://doi.org/10.1016/j.ijrobp.2009.06.091

Metlay, J. P., Waterer, G. W., Long, A. C., Anzueto, A., Brozek, J., Crothers, K., Cooley, L. A., Dean, N. C., Fine, M. J., Flanders, S. A., Griffin, M. R., Metersky, M. L., Musher, D. M., Restrepo, M. I., & Whitney, C. G. (2019). Diagnosis and treatment of adults with community-acquired pneumonia: An official clinical practice guideline of the American Thoracic Society and Infectious Diseases Society of America. American Journal of Respiratory and Critical Care Medicine, 200(7), e45–e67. https://doi.org/10.1164/rccm.201908-1581ST

Mishel, M. H. (1988). Uncertainty in illness. Image: The Journal of Nursing Scholarship, 20(4), 225–232. https://doi.org/10.1111/j.1547-5069.1988.tb00082.x

Naidoo, J., Wang, X., Woo, K. M., Iyriboz, T., Halpenny, D., Cunningham, J., Chaft, J. E., Segal, N. H., Callahan, M. K., Lesokhin, A. M., Rosenberg, J., Voss, M. H., Rudin, C. M., Rizvi, H., Hou, X., Rodriguez, K., Albano, M., Gordon, R.-A., Leduc, C., … Hellmann, M. D. (2017). Pneumonitis in patients treated with anti–programmed death-1/programmed death ligand 1 therapy. Journal of Clinical Oncology, 35(7), 709–717. https://doi.org/10.1200/JCO.2016.68.2005

Nishino, M., Giobbie-Hurder, A., Hatabu, H., Ramaiya, N. H., & Hodi, F. S. (2016). Incidence of programmed cell death 1 inhibitor–related pneumonitis in patients with advanced cancer: A systematic review and meta-analysis. JAMA Oncology, 2(12), 1607–1616. https://doi.org/10.1001/jamaoncol.2016.2453

Nosek, B. A., Ebersole, C. R., DeHaven, A. C., & Mellor, D. T. (2018). The preregistration revolution. Proceedings of the National Academy of Sciences, 115(11), 2600–2606. https://doi.org/10.1073/pnas.1708274114

Parshall, M. B., Schwartzstein, R. M., Adams, L., Banzett, R. B., Manning, H. L., Bourbeau, J., Calverley, P. M., Gift, A. G., Harver, A., Lareau, S. C., Mahler, D. A., Meek, P. M., & O’Donnell, D. E. (2012). An official American Thoracic Society statement: Update on the mechanisms, assessment, and management of dyspnea. American Journal of Respiratory and Critical Care Medicine, 185(4), 435–452. https://doi.org/10.1164/rccm.201111-2042ST

Prigerson, H. G., Bao, Y., Shah, M. A., Paulk, M. E., LeBlanc, T. W., Schneider, B. J., Garrido, M. M., Reid, M. C., Berlin, D. A., Adelson, K. B., Neugut, A. I., & Maciejewski, P. K. (2015). Chemotherapy use, performance status, and quality of life at the end of life. JAMA Oncology, 1(6), 778–784. https://doi.org/10.1001/jamaoncol.2015.2378

Puchalski, C., Ferrell, B., Virani, R., Otis-Green, S., Baird, P., Bull, J., Chochinov, H., Handzo, G., Nelson-Becker, H., Prince-Paul, M., Pugliese, K., & Sulmasy, D. (2009). Improving the quality of spiritual care as a dimension of palliative care: The report of the Consensus Conference. Journal of Palliative Medicine, 12(10), 885–904. https://doi.org/10.1089/jpm.2009.0142

Puxty, K., McLoone, P., Quasim, T., Kinsella, J., & Morrison, D. (2014). Survival in solid cancer patients following intensive care unit admission. Intensive Care Medicine, 40(10), 1409–1428. https://doi.org/10.1007/s00134-014-3471-9

Schillinger, D., Piette, J., Grumbach, K., Wang, F., Wilson, C., Daher, C., Leong-Grotz, K., Castro, C., & Bindman, A. B. (2003). Closing the loop: Physician communication with diabetic patients who have low health literacy. Archives of Internal Medicine, 163(1), 83–90. https://doi.org/10.1001/archinte.163.1.83

Singer, M., Deutschman, C. S., Seymour, C. W., Shankar-Hari, M., Annane, D., Bauer, M., Bellomo, R., Bernard, G. R., Chiche, J.-D., Coopersmith, C. M., Hotchkiss, R. S., Levy, M. M., Marshall, J. C., Martin, G. S., Opal, S. M., Rubenfeld, G. D., van der Poll, T., Vincent, J.-L., & Angus, D. C. (2016). The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA, 315(8), 801–810. https://doi.org/10.1001/jama.2016.0287

Skeoch, S., Weatherley, N., Swift, A. J., Oldroyd, A., Johns, C., Hayton, C., Giollo, A., Wild, J. M., Waterton, J. C., Buch, M., Linton, K., Bruce, I. N., Leonard, C., Bianchi, S., & Chaudhuri, N. (2018). Drug-induced interstitial lung disease: A systematic review. Journal of Clinical Medicine, 7(10), 356. https://doi.org/10.3390/jcm7100356

Tarantino, P., Modi, S., Tolaney, S. M., Cortés, J., Hamilton, E. P., Kim, S.-B., Toi, M., Andrè, F., & Curigliano, G. (2021). Interstitial lung disease induced by anti-ERBB2 antibody–drug conjugates: A review. JAMA Oncology, 7(12), 1873–1881. https://doi.org/10.1001/jamaoncol.2021.3595

Temel, J. S., Greer, J. A., Muzikansky, A., Gallagher, E. R., Admane, S., Jackson, V. A., Dahlin, C. M., Blinderman, C. D., Jacobsen, J., Pirl, W. F., Billings, J. A., & Lynch, T. J. (2010). Early palliative care for patients with metastatic non–small-cell lung cancer. New England Journal of Medicine, 363(8), 733–742. https://doi.org/10.1056/NEJMoa1000678

Teno, J. M., Clarridge, B. R., Casey, V., Welch, L. C., Wetle, T., Shield, R., & Mor, V. (2004). Family perspectives on end-of-life care at the last place of care. JAMA, 291(1), 88–93. https://doi.org/10.1001/jama.291.1.88

Tricco, A. C., Lillie, E., Zarin, W., O’Brien, K. K., Colquhoun, H., Levac, D., Moher, D., Peters, M. D. J., Horsley, T., Weeks, L., Hempel, S., Akl, E. A., Chang, C., McGowan, J., Stewart, L., Hartling, L., Aldcroft, A., Wilson, M. G., Garritty, C., … Straus, S. E. (2018). PRISMA extension for scoping reviews (PRISMA-ScR): Checklist and explanation. Annals of Internal Medicine, 169(7), 467–473. https://doi.org/10.7326/M18-0850

Vahid, B., & Marik, P. E. (2008). Pulmonary complications of novel antineoplastic agents for solid tumors. Chest, 133(2), 528–538. https://doi.org/10.1378/chest.07-0851

Wright, A. A., Zhang, B., Ray, A., Mack, J. W., Trice, E., Balboni, T., Mitchell, S. L., Jackson, V. A., Block, S. D., Maciejewski, P. K., & Prigerson, H. G. (2008). Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA, 300(14), 1665–1673. https://doi.org/10.1001/jama.2008.1525