|Home | About | Journals | Submit | Contact Us | Français|
People with ALS who choose tracheostomy demonstrate a strong and mostly consistent attachment to life from the point of diagnosis. It is unclear if these patients also use medical and health services to a greater degree than patients who decide against tracheostomy. In this research, patients with a high likelihood of dying over six months (forced vital capacity <50% predicted) were followed monthly until death or tracheostomy with long-term mechanical ventilation (LTMV). Patient service use was measured by caregiver reports of 1) ALS-specific prosthetic devices, 2) allied health or medical services, 3) legal preparation for medical care or the end of life, and 4) medical care episodes. Caregivers also reported all patient prescription medications. Over follow-up, 57 patients died and 14 elected to have tracheostomy and LTMV. Patients who opted for LTMV were younger and had higher household incomes. They were significantly more likely to use nasal ventilation, paid home care, and family or personal counseling over follow-up; and they were also more likely to remain on medications. The proactive orientation to health and desire to live despite severe disability reported for people choosing LTMV thus extends as well to more intensive use of medical and supportive care in the months prior to tracheostomy. A challenging task for clinicians is to acknowledge this strong desire to live while providing appropriate expectations for life after tracheostomy.
The choice of tracheostomy and long-term care in people with ALS remains relatively rare but shows wide variation across treatment sites (1.4% to 15% in the U.S.) (1) and countries (45% in Japan) (2). It may also be increasing. In a cohort followed across the last six months of life at one multidisciplinary ALS clinic, 18% opted for tracheostomy and long-term mechanical ventilation (3). Prior research has examined psychosocial and demographic correlates of this choice. Patients choosing tracheostomy are younger, more likely to have children in the home, less likely to report depressive symptoms, and more likely to demonstrate strong attachment to life at the point of diagnosis (4–6).
In this research, we asked if patients choosing tracheostomy also demonstrate greater medical and health service use relative to patients who decide against tracheostomy. Evidence of increased service use in the period prior to tracheostomy would indicate a consistent, continuing preference for treatment and prolongation of life in this subset of patients. Their orientation to disease management may accordingly differ from that of patients unwilling to consider tracheostomy. Patients who opt for tracheostomy may be more likely to seek alternative or experimental therapies, for example. They may also be more likely to report frustration with standard patient education and clinical management. In our experience, these patients seek more engagement by clinicians and report great frustration when they are told “go home, get your things in order, there is nothing you can do,” as one patient recently complained. This broad orientation may also lead to more strain within families, who must cope with disease progression and the prospect of post-tracheostomy care. For all these reasons, we sought to investigate whether patients opting for tracheostomy differ systematically in disease management over the course of disease progression.
Over 90% of patients were enrolled from the Eleanor & Lou Gehrig MDA/ALS Research Center at Columbia University. Other patients were referred from hospices, ALS support groups, or two other ALS multidisciplinary clinics. The clinic coordinator identified potential participants whose forced vital capacity was <50%, a value related to risk of hospice admission and death or need for mechanical ventilation within 6 months (7). Eligible patients did not meet criteria for dementia, spoke English, had an unpaid caregiver who was available for interview, were not using mechanical ventilation at baseline, were able to communicate at least “yes” and “no,” and lived within a 3-hour drive from the medical center. Study enrollment began in January 2000 and ended in June 2004. The Institutional Review Boards of Columbia-Presbyterian Medical Center and the New York State Psychiatric Institute approved the research protocol.
After the clinic coordinator identified potential participants, she described the study to patients and caregivers and obtained consent for the research team to make contact. The principal investigator then called for further explanation, answered questions about the study, and obtained verbal consent; informed consent from patients and caregivers was obtained during a home visit, when interviews were conducted. Interviews were scheduled at one-month intervals until patients met a study endpoint of tracheostomy or death. Patients and caregivers completed a separate interview on the same schedule. Caregivers also completed an interview after the death of patients.
Patient service use was measured by caregiver reports of 1) ALS-specific prosthetic devices (augmentative communication, feeding gastrostomy [PEG placement], nasal ventilation, suction device), 2) allied health or medical services (paid home care, personal or family counseling, alternative-complementary medicine, clinical trial participation, hospice), 3) legal preparation for medical care or the end of life (health care proxy, living will, power of attorney, autopsy consent), and 4) medical care episodes (emergency department, hospital, or skilled nursing facility admissions). This set of services captures the major options for disease management available to patients in later stages of ALS (8). Caregivers reported whether patients were using each service at monthly assessments.
Caregivers also reported patient prescription medications. They could report up to nine medications. Medications were categorized by drug indication, which yielded 26 medication categories. These were further reduced to four broad categories: 1) ALS-specific therapies (medications specifically designed to slow the progression of the disease: riluzole, high-dose vitamins, gabapentin, antioxidants, creatine, and clinical trial medications), 2) palliative medications (pharmacological therapies aimed at treating common symptoms of ALS, such as pain, sialorrhea, respiratory insufficiency, constipation, muscle spasms, cramping, and insomnia), 3) mood medications (medications aimed at treating depression, anxiety, agitation and emotional lability, such as alprazolam, buspirone, bupropion, paroxetine, and sertraline), and 4) “other” (medications to treat non-ALS chronic conditions and comorbidities).
Use of each service and medication type was noted at study baseline and at the last assessment before long-term mechanical ventilation or death. In addition, we calculated the proportion of patients who used each service or medication type across all assessments.
To assess severity of ALS, patients completed the ALS Functional Rating Scale (ALSFRS-R) (9) at each assessment. The ALSFRS consists of 12 items covering four domains: bulbar functions (speech, salivation, swallowing), fine motor skills (handwriting, handling utensils, dressing and hygiene), gross motor skills (turning in bed, walking, climbing stairs), and respiratory status (dyspnea, orthopnea, use of ventilation support). Each item is rated on a 5-point scale and a sum computed.
To assess depressive symptoms, interviewers administered the Patient Health Questionnaire (10). The PHQ assesses DSM-IV criteria for major and minor depressive disorder. While the Structured Clinical Interview for DSM-IV (SCID) is the gold standard diagnostic evaluation in psychiatry, it requires interactive queries and follow-up probes. A full diagnostic interview was not feasible in this sample of patients with end-stage ALS, where 40% were unable to speak at baseline (6). A clinical psychologist reviewed all PHQ assessments to assign DSM-IV depression diagnoses. Caregivers reported on patient cognition (memory, concentration, ability to follow a plot on television) and lability. Caregiver reports of patient cognitive status were validated against patient performance on the Mini-Mental State Exam (MMSE)(11). Patients whose caregivers reported memory problems performed more poorly on the MMSE, especially in word registration and recall.
Caregivers also reported level of burden and satisfaction associated with care. We used five items from the Zarit burden scale (12). These capture stress, not having time for oneself, and restriction in social and family life because of caregiving responsibilities. Satisfaction with caregiving commitment was assessed with an additional five items from Folkman’s measure of caregiving satisfaction (13).
Patients choosing tracheostomy and patients who died over follow-up were compared to determine if service use differed at baseline and over follow-up. Similar analyses were conducted for medication use. Proportions were compared using χ2 to test for differences. Because length of follow-up differed in the two patient groups, we used Mantel-Haenszel tests to stratify patients by number of study assessments (1, 2–4, 5+) and also developed logistic regression models to assess group differences adjusting for demographic indicators, study duration, and disease status. For assessing change between baseline and last assessments, we developed analysis of variance models using repeated measures.
Of 80 patients enrolled into the Living with ALS cohort, three patients did not have unpaid family or friend caregivers and were excluded from analyses. An additional six patients were excluded because they were lost to follow-up (n=2), did not die or elect to have long-term mechanical ventilation during the course of the study (n=2), or had caregivers who participated in a post-death interview alone (n=2). Excluding the nine left, 71 patients with caregivers who reported on service and medication use in the months before death or tracheostomy. Fifty-seven died and 14 elected to have long-term mechanical ventilation (LTMV). Two additional patients had unplanned emergency tracheostomies (one against her wishes) and died shortly after the procedure. These patients were not considered to have chosen tracheostomy.
Patients who died over follow-up had fewer days on study than patients who opted for tracheostomy and long-term mechanical ventilation (163 vs. 253 days, P = 0.13; median of 110 vs. 123 days). The median number of monthly assessments was three among patients who died and four among patients who opted for mechanical ventilation. The mean time from first assessment to tracheostomy/LTMV was 364 days and to death 207 days (P = 0.11). Sixty of the 77 patients completed more than one monthly assessment. Based on caregiver reports of the date of diagnosis, our first assessment took place an average of 20 months (17, median) after diagnosis.
At baseline, patients who died or chose mechanical ventilation differed in age (Table 1). Patients who later opted for LTMV were younger (51.4 vs. 64.5, P < 0.01). The two groups did not differ in other sociodemographic features, including gender, race, education, insurance, or presence of spousal caregiver. However, patients choosing LTMV had significantly higher income, and their caregivers were more likely to be employed (78.6% vs. 40.4%, P < 0.05) and complete college (85.7% vs. 56.1%, P < 0.05). Patients in the two groups were similar in level of disability, presence of cognitive symptoms, and time since diagnosis and symptom onset. Patients who later opted for LTMV were less likely to report symptoms consistent with major depression, but this difference did not achieve statistical significance.
Patients who later opted for LTMV were more likely to use medical and supportive care (Table 2). At baseline, 78.6% of the LTMV group used nasal ventilation, which increased to 92.9% across the follow-up period. In patients who died without LTMV, 51.8% used nasal ventilation at baseline and 62.5% over follow-up (P < 0.05). The LTMV group was also more likely to use supportive care, including personal or family counseling and paid home care. Use of feeding gastrostomies, augmentative communication devices, or alternative-complementary medicine did not significantly differ between the groups. Nor did the groups differ in legal preparation for medical care or the end of life, as indicated by the proportion with a health care proxy, living will, power of attorney, or autopsy consent.
Patients who opted for LTMV were less likely to be certified for hospice. At baseline, only one (7.1%) of the 14 patients who opted for LTMV used hospice, and another two patients used hospice at some point over follow-up, yielding a cumulative total of 21.4%. In contrast, among patients who did not choose LTMV, 43.9% were on hospice at baseline and 70.2% used hospice at some point over follow-up (P < 0.01).
The two groups did not differ in use of emergency department admissions, hospitalization, or skilled nursing home care.
Because the follow-up period was longer among LTMV patients, tests of differences in proportions between the two groups were adjusted for length of follow-up using Mantel-Haenszel tests. Differences remained significant in logistic regression models with duration of follow-up included as a covariate.
Looking only at the subset of people with more than one assessment over follow-up (47 deaths, 13 LTMV) confirmed these differences. In addition to greater use of nasal ventilation, counseling, and paid home care, patients opting for LTMV in this comparison were significantly more likely to report use of complementary/alternative therapies (61.5% vs. 27.7%, P = 0.02) and participation in clinical trials (53.8% vs. 23.4%, P = 0.03).
Patients choosing LTMV began follow-up with significantly less use of palliative medications (Table 3). They were also less likely to use “other” medications, that is, medications not defined as mood, palliative or ALS-specific therapies, which may be related to their younger age. Differences in medication use attenuated over follow-up.
Larger differences emerge when comparisons are restricted to the 60 patients (47 deaths, 13 LTMV) who were seen more than once over follow-up (Table 4). Patients who did not choose LTMV were more likely to stop medication use by their last study assessment before death. Significant time and group effects (P < 0.01) were confirmed in repeated measures analysis of variance models.
In this cohort, caregivers of patients who ultimately opted for tracheostomy reported greater depressive symptoms and higher burden at baseline than caregivers to patients who declined tracheostomy; however, burden and depressive symptoms declined in both groups over time (14). Service use in the two groups was not related to caregiver burden or satisfaction. Similarly, we did not observe a clear pattern of change in caregiver burden or satisfaction following tracheostomy. Of the 14 patients who began long-term mechanical ventilation, we were able to interview eight before and after patient tracheostomy. We found increases in burden in five, stability in two, and declines in one. Caregiver satisfaction was stable across the transition.
In this sample of ALS patients with late-stage disease, use of health services and medications was more likely in patients who ultimately opted for tracheostomy and LTMV. In the months prior to death or tracheostomy, patients who chose tracheostomy were more likely to use nasal ventilation, paid home care, and personal or family counseling. The two groups of patients did not differ in legal preparation for the end of life or medical care episodes. Not surprisingly, none of the patients who opted for LTMV gave consent for autopsy, while 16.7% of patients who died gave such consent. Also consistent with these outcomes, patients who died were more likely to use hospice. It is important to note, however, that three of the 14 patients who opted for LTMV were certified for hospice at some point over follow-up, suggesting that for some patients the choice of LTMV is less clear and comes late. The majority, however, appear to have a clear preference for tracheostomy early on, as also reported in a separate cohort of recently diagnosed patients, who were mostly enrolled within one year of diagnosis and followed for a median of 12 months (8). This research shows that patients choosing LTMV use medical and supportive care services consistent with this preference.
Patients choosing LTMV also had different medication profiles. They were less likely to use palliative medications. Also, unlike patients who were approaching death, patients opting for LTMV were likely to continue taking medication over the course of follow-up. Almost all patients who died stopped taking medications by the time of their last study assessment.
Comparing our late-stage ALS cohort to the more recently diagnosed cohort described above shows how service use increases with progression of the disease. The early-stage cohort had a mean forced vital capacity (FVC) of 78.5%. At baseline very few patients in the early-stage cohort used nasal ventilation (3.2%) or had PEG placement (1.1%) (8). By contrast, two-thirds of patients in the late-stage cohort used nasal ventilation and half had PEG placement at their initial assessment. Similar differences obtained for paid home care, counseling, and use of alternative-complementary medicine. The two cohorts are separated by five years, and changing practice patterns may explain some of these differences (15). Still, comparisons suggest that in the first 12 months after diagnosis patients are unlikely to use medical and supportive care services, but by 18–24 months half to two-thirds make use of advanced medical technologies (for example, nasal ventilation and PEG placement). Usage continues to increase in patients opting for LTMV but levels off and then declines in patients who decide against LTMV and approach death. For example, all but one patient who chose LTMV (92.9%) used nasal ventilation beforehand. The cumulative proportion among patients who did not choose LTMV was 62.5%.
Thus, the proactive orientation to health and desire to live despite severe disability reported for people choosing LTMV (3, 4) extends as well to more intensive use of medical and supportive care in the months prior to tracheostomy. Consistent with the strong preference to live, patients choosing LTMV also report fewer depressive symptoms in this period (3, 6). Yet outcomes following LTMV are poor. A recent Swedish series found that only 5% of people with ALS who choose LTMV survived five years (16). In an Italian series of LTMV ALS patients, 9% died in hospital and median survival among people discharged was 37 months (17). In our cohort, 4 (29%) patients on LTMV died over two years of follow-up, and half also reported a waning commitment to life after tracheostomy (3).
Limitations of this research include its reliance on patients mainly from a single tertiary care center who were willing to participate in a demanding interview schedule. This may limit generalizability of results and suggests a need for replication with other samples. Still, we have shown in prior research that our study cohort is similar to patients seen in the tertiary ALS care setting (3, 4). In addition, hospice eligibility in the U.S. requires that patients not pursue tracheostomy; the dynamics of service use in ALS might be different in health systems that do not require such a choice.
These results suggest that patients seeking tracheostomy have a different orientation to disease management than patients unwilling to consider tracheostomy. A challenging task for clinicians is to acknowledge the strong desire to live among this subset of patients while providing appropriate expectations for life after tracheostomy.
This research was supported by National Institute of Mental Health grant R01 MH62200.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.