Physical Examination, Complementary Tests, and Functional Status Assessment

In a critically ill patient with malignant central airway obstruction (CAO) who requires ventilatory assistance, interventional bronchoscopic procedures may be lifesaving, may allow extubation and time for additional therapies to be initiated, and may prolong survival.^1–6 Reported overall median survival rates of critically ill patients with malignant CAO vary between 2 and 38 months post bronchoscopic interventions,^1,6,7 and 1 year survival reportedly has been as low as 15%.⁸ This wide variability in results is probably explained by inclusion of patients at various stages of their disease, with those with various types of malignancies undergoing various treatments before and after interventions, and receiving different degrees of respiratory support. These critically ill patients are often admitted emergently to the intensive care unit (ICU) only when respiratory failure develops. According to Society of Critical Care Medicine (SCCM) guidelines on ICU triage, admissions, and discharge, patients with advanced cancer should be admitted to the ICU when a reversible cause (e.g., pulmonary embolism, tamponade, airway obstruction) is identified.* Patients with advanced malignancy complicated by airway obstruction are given priority 3, which is defined as “unstable patients who are critically ill but have a reduced likelihood of recovery because of underlying disease or nature of their acute illness.”⁹ According to these guidelines, patients “may receive intensive treatment to relieve acute illness, however, limits on therapeutic efforts may be set, such as no intubation or cardiopulmonary resuscitation.” In practice, patients with acute respiratory failure usually are intubated, placed on mechanical ventilation, and transferred to the ICU. Patients with non–small cell lung cancer (NSCLC) and respiratory failure from CAO, such as our patient, therefore may warrant invasive mechanical ventilatory support and admission to the ICU, pending evaluation for palliative bronchoscopic interventions.¹⁰ If the bronchoscopic intervention is successful and the patient’s physiologic status has stabilized, ICU monitoring and care may no longer be necessary, and the patient can be transferred to a unit that provides a lower level of care. If, on the other hand, airway interventions are not successful, or if a patient’s physiologic status has deteriorated and active interventions are no longer planned, discharge to a lower level of care, including withdrawal of life support or referral to hospice, is a reasonable consideration. Families and referring physicians can be comforted by knowing that bronchoscopic intervention was not indicated, or that is was performed but was unsuccessful in alleviating the airway obstruction.

Many patients and their families may wonder whether all possible therapeutic alternatives are being considered, especially when the patient is possibly in the terminal stages of disease. Truthful disclosure by physicians is crucial in these instances to avoid compromising a physician-patient relationship built on trust. Although some patients and families are averse to truthfulness, most want their doctors to be honest and forthcoming in their communications about diagnosis and prognosis, even when it means announcing the likelihood of death. In addition to other factors, such as age and disease diagnosis, ethnicity may play an important role in a patient’s perceptions regarding truth telling and availability of therapeutic interventions. For example, approximately one in five deaths in the United States occurs in or shortly after a stay in the ICU,¹¹ where evidence of disparities in end-of-life care has been found. In the United States, racial minorities apparently have been subjected to lower quality of care than whites,¹² and although these findings are not consistent,*^13,14 some studies report that African American patients receive fewer medical interventions, have shorter lengths of stay, and use fewer resources. One large study, however, evaluated the medical charts of patients who died in 15 intensive care units (3138 patients, of whom 2479 [79%] were white and 659 [21%] were nonwhite [e.g., African American, Asian, Pacific Islander, American Indian, Hispanic]). Logistic regression adjusted for socioeconomic factors measured by education, income, and insurance status showed that nonwhite patients were less likely to have living wills, were more likely to die with full support including cardiopulmonary resuscitation (CPR), and were more likely to die in a setting of full support than were white patients. In addition, nonwhites were more likely to have undergone medical interventions during their ICU stay, including dialysis, vasopressors, and mechanical ventilation.*¹⁵ Although all patients are and should be treated as VIPs (very important persons), consideration and respect for patient and family perceptions regarding therapy and truth telling are warranted, particularly if factors known to affect perceptions and behaviors are present.

The patient’s laboratory tests revealed hypoalbuminemia and moderate hyponatremia. The latter was likely due to the syndrome of inappropriate antidiuretic hormone (SIADH) secretion†¹⁶ because the patient was euvolemic, and no signs of hypovolemic hyponatremia, such as gastrointestinal losses, excessive diuresis or adrenal insufficiency, or salt wasting nephropathy or cerebral salt wasting, were noted. Other laboratory tests revealed a blood urea nitrogen (BUN) level less than 10 mg/dL and urine osmolarity greater than 100 mOsm/kg of water—findings consistent with the diagnosis of SIADH.

Comorbidities

This patient had chronic obstructive pulmonary disease (COPD) and atrial fibrillation, increasing the risks for postoperative pulmonary and cardiovascular complications, respectively, in cases of interventions performed under general anesthesia.

Support System

The patient had no medical insurance. The family expressed their frustration that before this event of respiratory failure, the patient’s access to care had been very limited and, they believed, suboptimal; they mentioned to us that urgent care centers on two occasions prescribed a 5 day antibiotic course but had not performed imaging studies or other tests, even though health care providers had been told about his prior history of lung cancer. One hates to think that differences in insurance status might contribute directly to different outcomes of patients suffering from lung cancer. Surely, few patients understand the complexities of the health insurance system in the United States, where copayments of various types, a mixture of health maintenance organization/preferred provider organization (HMO/PPO) plans, and constantly changing rules regarding deductibles and the need for authorizations can trouble the consumer, especially when illness strikes. Few patients understand that virtually all health plans provide for coverage in case of an emergency, regardless of geographic location, and that patients have a right to request and advocate for laboratory tests and imaging studies if they believe these will be important for their care.

In addition, differences in race, ethnicity, income, and education may have an effect on processes of care and outcomes. In the United States, results of studies show that patients with Medicaid* or no insurance had worse outcomes than other patients suffering from lung cancer. Although some of these disparities may have been secondary to confounding factors such as smoking and other health behaviors, available data suggest that patients with lung cancer without insurance do poorly because access to care is limited, and/or because they present with more advanced disease that is less amenable to treatment.¹⁷ On the other hand, the risk for Medicare^† patients has been found to be similar to that for patients with private insurance; for those with Medicare disability insurance, no differences in stage were noted when a fee-for-service status was compared with a health maintenance organization or combination insurance status.¹⁸

Patient Preferences and Expectations

This patient with a known lung cancer had health care advance directives with regard to a Do Not Resuscitate (DNR) order in case of terminal irreversible disease. In general, published evidence reveals a low rate of completion of these documents among Hispanics.¹⁹ In addition to lack of trust in the health care system, health care disparities, and cultural perspectives on death and suffering, lack of acceptance of advance directives among Hispanics may originate from a view of collective family responsibility.²⁰ In many Latin American countries, in fact, illness is viewed as a family affair rather than as a struggle of individual suffering. Before his respiratory failure, our patient had expressed his wish to not be on “breathing machines” if the disease process became irreversible. Therefore a DNR had been placed in his medical record upon his admission to the outside facility. One might recall that cardiopulmonary resuscitation (CPR) was originally designed for the prevention of sudden, unexpected death—not necessarily to prolong the life of an irreversibly terminally ill person.²¹ At the same time, the quality and duration of life-related value judgments should be avoided, and sanctity of life arguments can be made to justify CPR in virtually any setting.

In making decisions about code status orders, physicians and patients must communicate effectively, so that patients can receive informed and compassionate care that respects their wishes. Communication between physicians and patients (or their surrogates) about code status orders, however, is difficult, especially if matters are being discussed emergently at the time of a medical crisis. Misunderstandings about code status and overall goals of care preferences may lead to unwanted medical interventions or withholding/withdrawing of desired interventions. Results of studies show that patients in the ICU and their surrogates have insufficient knowledge about in-hospital CPR and its likelihood of success, such as whether CPR will result in preservation of life, preservation of organ function, discharge from an intensive care unit, or discharge alive from the hospital. In general, it appears that many patients, families, and physicians probably share an excessively positive estimation of the outcomes of CPR.²²

A patient’s code status preferences may not always be reflected in code status orders, and assessments may differ between patients, their surrogates, family members, and physicians about what goal of care is most important.²² In our case, the patient’s family was particularly concerned that any de-escalation of the patient’s care, including proceeding with extubation or removal from mechanical ventilation, would constitute euthanasia. They also expressed firm disagreement with a proposed strategy of palliative sedation.* Oftentimes, when disagreements occur, an ethics consultation can help clarify terminology and avoid adversarial encounters between the treating team and the family. Anyone may request the assistance of ethics consultants, including family members and even health care providers not involved in the care of a particular patient if they sense impending difficulties in the medical encounters or in their own dealings with conflict. The involvement of other disciplines, including palliative care medicine, chaplaincy, and a social worker to address the family’s concerns, is a reasonable first step. In our case, physicians explained to the family that euthanasia is completely different from palliative sedation or withdrawal of life-sustaining treatment based on its intent and outcome. In euthanasia, the intent is deliberate termination of a patient’s life (which is illegal in the United States), and palliative sedation is designed to relieve symptoms or the effects of burdensome interventions (i.e., mechanical ventilation). This is ethically justified and legal, even when sedation might cause, from its double effect, a hastened demise.²³

Indications

Relieving central airway obstruction could prevent post obstructive pneumonia, sepsis, and septic shock; allow extubation and a change in the level of care; permit initiation of systemic therapy; and potentially improve survival. Evidence suggests that bronchoscopic therapies often provide acute relief of the obstruction, improve quality of life, and serve as a therapeutic bridge until systemic treatments become effective.^1,4,7 Subsequent systemic treatments (chemotherapy and/or radiotherapy) were shown to increase disease-free survival during the first year after restoration of airway patency.^1,8

In patients with CAO (lobar or mainstem), assessing the functionality of the lung parenchyma distal to the obstruction is useful when interventions meant to establish airway patency are considered. Functionality of the lung distal to the obstruction may not be restored in patients who have had chronic complete obstruction and lack of ventilation (Figure 26-2). Determining whether there is functional airway and lung beyond an obstruction is essential for any successful bronchoscopic intervention,* in part because significant friability of bleeding from thin infiltrated bronchial mucosa, or lack of lung perfusion despite restored airway patency, might preclude intervention. In one study, 71% of patients who initiated radiation therapy within 2 weeks after radiologic evidence of atelectasis had complete re-expansion of their lungs, compared with only 23% of those irradiated after 2 weeks.²⁴ Studies pertaining to successful bronchoscopic treatment and time to treatment are lacking.

Figure 26-2 Images obtained from a different patient show the right lower lobe and middle lobe atelectasis (A) on chest radiograph and (B) on chest computed tomography, as well as (C) complete closure of the bronchus intermedius by a mucosal infiltrating tumor.

One way to assess the perfusion status of lung parenchyma distal to an airway obstruction is to attempt bypass of the stenosis using a high-resolution endobronchial ultrasonography (EBUS) radial probe. For instance, the perfusion to one lobe could be completely shut down in a case of complete bronchial obstruction simply because of the Euler-Liljestrand reflex,^† not because of pulmonary artery involvement. In cases of frank pulmonary artery occlusion, however, establishing airway patency could result in increased dead space ventilation. This finding should probably annul further intervention.²⁵ Our patient had a critical distal left main bronchial obstruction in the setting of right pneumonectomy, but lobar and segmental airways in the left lung were patent bronchoscopically, and distal lung parenchyma appeared functional on the chest radiograph (see Figure 26-1).

Contraindications

No absolute contraindications to bronchoscopic interventions performed under moderate sedation or general anesthesia were identified.

Expected Results

Patients with advanced NSCLC and CAO who undergo successful interventional bronchoscopy to relieve airway obstruction might have a survival similar to those without CAO.²⁶ Studies also show that patients with respiratory failure and malignant CAO palliated by bronchoscopic intervention who underwent additional definitive therapy survived longer (median, 38.2 months; range, 1.7 to 57.0 months) than those who did not (median, 6.2 months; range, 0.1 to 33.7 months; P < .001).⁷ Successful removal from mechanical ventilation in patients with CAO due to malignancy has been reported in 50% to 100% of patients in several small case series^2,4 and, in some, prolonged survival was documented when systemic treatment was also administered (98 vs. 8.5 days).¹

Team Experience

Rigid bronchoscopic interventions with laser and/or stent placement are performed on a weekly basis at this institution by a dedicated interventional bronchoscopy team with enough expertise and experience to perform a procedure in a patient in poor general condition and with only one insufficiently ventilated lung.

Therapeutic Alternatives

Withdrawal of mechanical ventilation and initiation of palliative sedation can be considered in patients with respiratory failure requiring mechanical ventilation in the setting of advanced malignant CAO and for whom no additional systemic or local (including bronchoscopy) therapies are indicated or feasible. These interventions, however, are used in patients who warrant aggressive symptom control and still have severe suffering (most commonly physical symptoms such as delirium, dyspnea, and pain) related to their underlying disease or therapy-related adverse effects.²³ The effectiveness* of palliative sedation is in the range of 70% to 90%. In our patient, we decided to attempt relief of the airway obstruction using bronchoscopic techniques; had this been unsuccessful, we would have consulted palliative care medicine and discussed with the patient’s family withdrawal of life-sustaining treatments.^23†

Cost-Effectiveness

The costs of interventional bronchoscopic procedures such as laser resection and stent insertion are probably small compared with the costs of repeated external beam radiation therapy, palliative chemotherapy, or prolonged ICU stays for already advanced and refractory disease. These economic factors are obviously institution, insurer, and country specific.

Informed Consent

The patient was sedated and was on mechanical ventilation. We discussed the indications, risks, benefits, and alternatives of bronchoscopic intervention with the patient’s wife and family during ICU family meetings. Using certified interpreters, we assessed their understanding of the proposed interventions and options—essential for an autonomous decision. Although it may seem paradoxical, patients and their surrogates benefit from the help of others to make truly autonomous decisions because they require accurate and comprehensive information. They may also need someone to help them relate the information they have to what they care about. Furthermore, because most people do not have a clear rank of priorities ready at hand for all situations, most need help identifying what they care about most and how this relates to the current health care encounter.²⁷

Anesthesia and Perioperative Care

In the perioperative setting, anesthesiologists and surgeons may need to reconsider and re-evaluate standing DNR orders, which may need to be temporarily suspended during the intraoperative and immediate postoperative periods. Anesthesiologists may be skeptical of what they read in the patient’s chart, especially because it is possible that up until the time of the procedure, they usually know very little about the patient.* The surgeon may not be the patient’s primary physician and may not know the details of the circumstances in which the DNR decision was made. It is nearly impossible for advance directives to address adequately the multitude of clinical situations that may be encountered during the operative setting. CPR, if needed during the operative and immediate postoperative periods, includes not only chest compressions or electrical shocks, but also what could be viewed as excessive vasopressor usage or an inappropriate duration of resuscitation. Therefore it is important that all such patients be seen by the anesthesiologist and the surgeon preoperatively, and that a strategy be mutually agreed upon in case of the patient’s “death.” Although logistically difficult, a multidisciplinary discussion is warranted and might include the patient and/or the patient’s representative surrogate, the anesthesiologist, the surgeon, and the intensivist, who may be called upon postoperatively.²⁸ A palliative care specialist or medical ethicist can also provide helpful insights. Multidisciplinary discussions may allow time to address the complex ethical and practical issues often present in these situations. For example, in the United States, patients who do not speak English fluently generally are less likely than more fluent English speakers to be actively engaged in end-of-life decision making, even in less hurried situations.²⁹ Language barriers can be considerable, and working through translators, although necessary, is not always easy. Perioperative DNR orders also raise many issues of special concern. The DNR discussion with the patient or appropriate surrogate should include information about the characteristics of a possible resuscitation during the perioperative period; the risks, the benefits, and the likely outcome; and the reasons why resuscitation during the perioperative period might be determined by the patient, the surrogate, or the health care provider to be more or less burdensome than beneficial. If a surrogate is making the decision, the medical record should note the surrogate’s relationship to the patient (e.g., health care agent, spouse, guardian) and the basis on which the surrogate’s decision is made (e.g., “patient’s prior wishes,” “best interests”).

The DNR order should be written in detail and signed by one of the treating physicians. If the DNR order is cancelled preoperatively, as is often the case, then the time period and circumstances under which it is to be re-enacted should be specified.³⁰ Interviews with terminally ill patients about perioperative resuscitative orders have revealed that some patients wanted their preoperative DNR orders revoked, some wanted to use procedure-directed perioperative DNR orders, and others wanted to redefine their goals for the procedure and postoperative outcome. This requires that the anesthesiologist and the surgeon decide about appropriate means for resuscitation during the procedure,³¹ and certain limits may be set. For example, we have seen patients with advanced illness having a respiratory arrest at the end of a procedure, who are responsive to intubation and ventilation but who never require chest percussion or vasopressors. We have also seen cardiac arrest rapidly reversible by defibrillation and not requiring reintubation or pressors. Special attention should be paid to issues related to the response to any iatrogenic arrest.³² Surrogates can be good decision makers for others when they know specifically what the person would have wanted, or when they have a strong general impression of that person’s values and beliefs. In some cases, however, surrogate decision making will be an approximation, subject to conscious and unconscious biases, concerns, and conflicts; in studies assessing agreement between a patient and surrogate regarding the use of CPR, the percent of agreement ranged from only 53% to 90%, with the frequency of agreement least in sicker patients.³³

With regard to our patient’s preoperative physiologic optimization, fluid restriction (<1000 mL/day) was initiated during the ICU stay. Twenty-four hours later, his sodium had improved from 128 to 132 mEq/L. Had this not shown improvement, demeclocycline (300 to 600 mg orally twice daily), conivaptan (20 to 40 mg/day intravenously), tolvaptan (10 to 60 mg/day orally), or even hypertonic (3%) saline at less than 1 to 2 mL/kg/hr¹⁶* could have been used to correct hyponatremia preoperatively.