Abstract
Objective: To determine whether ALS patients’ preferences for ameliorative or life-extending technologies elicited early in the disease were related to later treatment choices.
Methods: In this prospective cohort study, 121 patients were seen at a tertiary ALS care center and followed for a median of 12 months. At baseline, patient preferences for use of tracheostomy and percutaneous endoscopic gastrostomy (PEG) placement were elicited. All patients received the same educational information before being interviewed about treatment preferences. Patients were then followed to determine if patients who viewed the interventions favorably at baseline were significantly more likely to use the interventions over follow-up.
Results: Six to twelve percent of patients were certain they wanted tracheostomy and 28.2% wanted PEG. Preferences were related to later treatment choices: 20% of patients who found tracheostomy acceptable had one in the follow-up period, compared with 3.4% of those not in favor (p < 0.001). For PEG, similar findings were obtained: 48.5% who initially found it acceptable had PEG, versus 8.1% of those not in favor of this treatment (p < 0.001). Patients who found the interventions acceptable were more likely to be recently diagnosed, expressed greater attachment to life, and showed greater declines in pulmonary function over follow-up.
Conclusions: Patients with ALS were able to express their preferences for life-extending or ameliorative technologies and made choices consistent with these preferences. However, patient preferences may change over time, and clinical education efforts are required throughout the course of disease.
The inescapably clear course and short survival time of ALS forces patients, their families, and their treatment teams to consider the acceptability of life-extending or ameliorative technology. Some studies have found a survival benefit for gastrostomy,1 bi-level positive airway pressure ventilation (BiPAP),2 and tracheostomy with mechanical ventilation,3 although neurologists may suggest these treatments more for their quality of life benefit. Some authors found patient satisfaction with mechanical ventilation,4,5 but the studies were necessarily restricted to patients who used the technology. Other studies show that patients may choose to discontinue use of such technologies,4,6 indicating an unacceptable quality of life, problems for caregivers, or unsustainable cost. We do not know the perspective of patients who decline to use these technologies.
We investigated these issues in a prospective cohort study of recently diagnosed ALS patients. We interviewed patients to elicit preferences for use of tracheostomy and percutaneous endoscopic gastrostomy (PEG) placement. Patients were then followed to examine how well these preferences predicted actual use of technology later. We are unaware of any other prospective study that assessed the extent to which patients can state preferences and also examined how well these preferences predicted actual treatment decisions.
Sample and methods.
In 1996 through 1997, 121 ALS patients from the Eleanor and Lou Gehrig MDA/ALS Center at Columbia-Presbyterian Medical Center, New York, were enrolled into a prospective study designed to examine palliative care decision making. Sixty-one percent of eligible patients agreed to participate. Patients did not participate because of refusal (25.8%), clinician judgments that the interview was inappropriate because of a patient’s psychiatric status (7.1%), or unwillingness of treating physicians to refer patients to the study (6.1%). Patients entering the cohort did not differ from nonparticipants in sociodemographic characteristics or features of the disease. Consistent with our goal of enrolling recently diagnosed patients, half of the cohort was enrolled within 3 months of diagnosis. A quarter of the sample had been diagnosed with ALS 12 or more months before the baseline interview.
The baseline interview for the cohort was conducted after patients had already spoken to their treating physician about the diagnosis of ALS, and after a nurse specializing in ALS had provided all patients with a standardized packet of educational materials about treatment choices in ALS and provided counseling about disease course and care. The standardized clinical education materials covered tracheostomy, PEG placement, and noninvasive nasal ventilation, as well as other issues such as autopsy, advance care directives, and augmentative communication technologies. The educational materials (available upon request) were factual and detailed and always stressed patient choice; for example, information on tracheostomy stated that “the manner in which each individual approaches the stage of illness and available choices is unique for each individual.”
Patients were eligible for the study if they met criteria for ALS, as indicated by finding all of the following signs: focal weakness, wasting, and fasciculations; upper motor neuron signs (increased tendon jerks, Hoffmann and Babinski signs, clonus); electromyographic evidence of denervation in three limbs; and no other identifiable cause for these symptoms. In essence, they met El Escorial criteria for probable or definite ALS. In addition, patients had to be followed at Columbia-Presbyterian Medical Center and speak English or a language for which a medical translator was available (Russian, Spanish; n = 3).
At baseline, respondents were interviewed on the same day as their clinical evaluation at the MDA/ALS clinic. Interviews lasted 1 hour and were conducted by trained researchers who were not providing clinical care. Information was also obtained about sociodemographic status, mental health, attachment to life, and health-related quality of life. Because of time limitations, approximately 15% of patients completed the questionnaire at home and returned it by mail. When dysarthria prevented patients from speaking (n = 8), they were asked to point to answers in the structured questionnaire. When patients could not speak or use hands (n = 4), they indicated responses with eye movements.
Once enrolled, patients were called or seen every 4 months to determine if prespecified endpoints had been reached. Major endpoints considered here include use of PEG and tracheostomy, but the entire spectrum of palliative care options was assessed, including use of nasal ventilation (BiPAP), adaptive aids and adjuvant therapies, hospice, advance care directives, psychiatric consultation, and counseling and support group participation. In every case, dates of first use were obtained to give the clearest picture of the timing of palliative care in ALS.
Each patient’s treating physician was also asked to complete a brief survey on what he or she typically tells patients after a diagnosis of ALS. These survey responses were anonymous (and, accordingly, could not be linked to information from particular patients). Seventy of the 95 physicians (73.4%) approached completed the mailed survey.
The research was approved by the Columbia-Presbyterian Medical Center Institutional Review Board, and informed consent was obtained from all subjects.
Measures.
Preferences for use of interventions.
Preferences for tracheostomy and PEG, as well as use of nasal ventilation (BiPAP), were obtained in two ways. In one series of questions, we asked patients how certain they were that they would want a treatment when their condition might warrant the intervention. Patients could endorse one of five options: absolutely against, uncertain but leaning against, undecided, uncertain but leaning toward, or absolutely want. We dichotomized these responses to identify a group leaning toward or absolutely in favor of the treatment and a group less in favor. We also noted patients who were unwilling to answer the questions, retaining them in analyses.
A second series of questions asked patients to indicate the longest time they would want to live with the intervention. Patients could endorse one of four options: would not want to live, willing to live 1 to 6 months, willing to live 7 to 12 months, or would want to live as long as possible. Here, too, we dichotomized responses, identifying a group willing to live ≥6 months or as long as possible with the technology and a group unwilling to live with the technology or willing to live with it for a period under 6 months. Two different questions were asked because we wanted to assess the consistency of patient responses across different question formats.
Factors related to patient preferences.
To understand factors related to patient preferences, we examined domains that might distinguish between patients willing to use the technologies and those who view them as unacceptable. We assessed a number of such domains, including patient mental health and attachment to life, as well as disease severity at the baseline interview. We also determined if patients had designated a health care proxy.
Patient mental health.
Patient mental health was assessed with the Beck Depression Inventory,7 a validated self-report instrument that is considered the gold standard of self-report scales intended to measure the severity of depressive symptoms. Future outlook was assessed with the Beck Hopelessness Scale, a 20-item self-report measure.8 Both measures have been shown to be reliable in a wide range of patient populations.
Attachment to life.
We assessed attachment to life by asking patients if they had a strong will to live, found meaning in their lives, had something to look forward to each day, expected to accomplish long-term goals, continued to look to the future, and would be bothered if their lives ended soon. Patients were asked to indicate how much they agreed or disagreed with these six statements on a five-point Likert scale. Coefficient alpha for this newly developed measure was 0.74, indicating acceptable internal consistency. The measure was designed to try to assess a patient’s desire to live despite potentially poor quality of life.
Severity of ALS.
Severity of ALS was assessed in a number of ways. All patients rated the extent to which the disease had affected a series of quality-of-life domains, including physical comfort, mobility, ability to work or do chores, and ability to perform activities of daily living (e.g., wash, bathe, use toilet). They indicated whether the disease had a major negative effect, some negative effect, or little negative effect on a five-point rating scale. In addition, 75 patients completed pulmonary function testing (forced vital capacity [FVC], expressed as a percentage of predicted capacity based on gender and age). A partly overlapping subset of patients (n = 72) was assessed by the nurse with the Appel scale.9 The scale provides ratings of the impact of ALS on swallowing and respiration, as well as other domains. Patients receiving pulmonary or nursing assessments did not differ from those who did not receive the assessments. We also obtained information on the duration of disease, expressed as the difference between the date of the first reported symptom and baseline interview, or, alternatively, between the date of diagnosis and baseline interview.
Analyses.
The relationship between patient preferences at baseline and actual use over the course of follow-up was examined in Kaplan-Meier survival analyses. Patients opposed to and in favor of each technology were compared according to time to first use of the technology; that is, the number of months on study until use of the technology, last assessment date, or time of death. Survival models were developed for each endpoint, and time-to-onset distributions were compared using the log-rank test.
A secondary aim of the study was to determine how patients who expressed a desire to use ameliorative or life-extending technologies may have differed from those who did not. Here we asked whether patients preferring or rejecting technologies differed on mental health measures, attachment to life, and indicators of disease severity. For these analyses, we compared the two groups at baseline by examining differences in mean and median values.
Results.
The mean age of this cohort of ALS patients was 59.7 (SD 12.7) years, with a range of 29 to 87 years. Sixty-one percent were men and 82% were white. Nearly 25% were born outside the United States. The cohort completed a mean of 14.2 years of school (SD 3.4). The median annual income was $35,000. All but one patient were living at home at the baseline assessment; one third were still employed.
FVC at baseline ranged from 21% of predicted to 118%; the median was 78% of predicted value. A quarter of the patients reported severe mobility problems, a third severe difficulty in the ability to perform daily chores, and 15% a need for assistance in activities of daily living. By Appel ratings, 58.3% of patients had normal respiratory function, 26.4% mild difficulty, and 15.3% more severe difficulty. A total of 61.1% were eating a general diet, 25% a soft diet, and 13.9% a mechanical soft diet or more restrictive diet.
To date, 60% of the cohort has been followed for 12 or more months. Over the 2.5 years since we first enrolled patients, 49% (59/121) of patients have died. The median survival time for the entire cohort using Kaplan-Meier estimates (which take into account differential follow-up) was 16 months.
At baseline, two patients already had a tracheostomy; five, PEG; and four, BiPAP. These patients were excluded from the analysis of time to these endpoints. There were 30 cases of incident PEG use (25.9%, 30/116), 8 incident tracheostomies (6.7%, 8/119), and 19 incident uses of BiPAP (16.2%, 19/117) in the follow-up period. With 24 months of follow-up, the cumulative incidence of tracheostomy was 20%, PEG 40%, and BiPAP 33%, as estimated by Kaplan-Meier analyses. PEG use and tracheostomy were highly associated: of 30 patients with PEG placement, 7 (23.3%) had tracheostomy, compared to only 3 of 91 patients (3.3%) without PEG (p = 0.001). Tracheostomy and BiPAP were less strongly related: of 23 patients with BiPAP, 4 (17.4%) had tracheostomy, compared to 6 of 98 patients (6.1%) without BiPAP (p = 0.08).
At baseline, 31.5% of patients had completed a health care proxy form.
Preferences for tracheostomy.
The table is a cross-classification of responses to the two questions we used to establish preferences for tracheostomy. The table summarizes responses of 119 subjects because it excludes the two subjects already on tracheostomy at baseline (who both stated that they wanted to live as long as possible with tracheostomy). Over 90% of the cohort answered the questions, indicating strong face validity (9.2% [n = 11] did not answer both questions). Sixteen subjects were undecided about tracheostomy on the attitude measure, but only two did not answer the question on the longest time they would want to live with a tracheostomy. The association between the two measures was significant (p < 0.0001 in analyses excluding subjects unable to answer either item). ⇓
Preferences of ALS patients for tracheostomy/mechanical ventilation: Cross-classification of responses
Using this approach, 5.9% (7/119) of the cohort were clear advocates for tracheostomy; they were “absolutely certain” they would choose tracheostomy and said they wanted to live as long as possible with tracheostomy. A total of 12.6% (15/119) could be considered advocates based on either of the measures (“leaning toward/undecided” or “absolutely want,” or willing to live >6 months). A total of 32.7% (39/119) rejected tracheostomy on any terms; they reported they were absolutely against it and would not want to live with the technology for any length of time. About 55% of patients were neither strongly opposed to or in favor of the technology.
Of 15 patients strongly in favor of tracheostomy (defined as wanting to live ≥6 months on tracheostomy), 20% (3/15) had tracheostomy; by contrast, of the 87 subjects opposed to tracheostomy, 3.4% (3/87) had tracheostomy. Seventeen subjects declined to answer the question; 11.8% (2/17) of these patients had tracheostomy. Time to tracheostomy was significantly earlier among subjects who favored tracheostomy, compared to those opposed (p < 0.03). Those who declined to answer the question were also more likely to have tracheostomy, although differences were not significant (figure 1). Patients completing a health care proxy form were less likely to view tracheostomy favorably (p < 0.02).
Figure 1. Baseline preference for tracheostomy and incidence of tracheostomy. In favor = willing to live >6 months with tracheostomy. Significantly earlier onset of tracheostomy compared with opposed group, p < 0.03.
Preferences for PEG placement.
A total of 28.2% reported they were in favor of PEG placement (these patients were “leaning toward/undecided” or “absolutely certain” they wanted PEG placement). Those in favor of PEG at baseline and those unable to specify a preference were significantly more likely to undergo PEG placement in the follow-up period compared with those opposed (figure 2). Excluding subjects with PEG at baseline, the proportion of patients who had PEG in the follow-up period was 48.5% among those in favor, 40.0% among those unable to answer the question, and 8.1% among those opposed. Time to onset of PEG was significantly earlier among subjects in favor of or unable to specify a preference (p < 0.001 for both comparisons).
Figure 2. Baseline preference for percutaneous endoscopic gastrostomy (PEG) and incidence of PEG. Incidence of PEG significantly different between opposed and other groups, p < 0.001.
Preferences for tracheostomy and PEG were highly correlated (r = 0.62, p < 0.0001). Very few patients endorsed tracheostomy without also endorsing PEG (<2% of the sample). Conversely, nearly 20% reported a preference for PEG while rejecting tracheostomy. Completion of a health care proxy and PEG preference were not related.
Preferences for BiPAP use.
We also examined preferences and outcomes involving noninvasive respiratory support (BiPAP) because this treatment is a less extreme intervention in which the relationship between preference and later use might be less pronounced. Excluding prevalent cases of BiPAP use, we found that 30.2% of subjects were in favor of BiPAP (“leaning toward/undecided” or “absolutely certain” they wanted BiPAP if their condition warranted it). Here, too, baseline preferences were strongly associated with use of BiPAP later. Time to first use was significantly earlier among this group than among those opposed (p < 0.02).
Correlates of patient preferences.
Preferences for use of the technologies were unrelated to age, education, or gender.
Patients in favor of tracheostomy or PEG placement at baseline did not have significantly more advanced disease than those opposed to use of the technologies. Patients did not differ in the self-rated impact of ALS on quality of life domains, in baseline FVC, or in nursing evaluations of respiration. Patients in favor of PEG also did not significantly differ in these indicators of disease progression or in nursing evaluations of swallowing, although swallowing was more severely affected among patients in favor of PEG (median Appel value of 6, “soft diet,” versus 3, “normal diet,” among those not in favor of PEG). Preferences for PEG were also associated with more rapidly progressing disease. Patients in favor of PEG at baseline were likely to have lower FVC ratings at last follow-up (mean FVC of 56% versus 71% among those opposed to PEG, p < 0.02).
However, longer duration of disease was associated with a lower preference for use of PEG and tracheostomy. Among patients diagnosed over a year before the interview, 19.0% were in favor of PEG; among patients diagnosed less than 3 months before the interview, 38.3% were in favor. For tracheostomy, similar differences were observed: 4.8% versus 22.7%, respectively. These differences approached significance (p < 0.10). Similar findings were observed when duration of disease was based on first reported symptoms.
Preferences were not related to mental health indicators (Beck Depression and Hopelessness scales). By contrast, preferences were significantly associated with our measure of attachment to life. Patients in favor of the technologies scored higher on this measure (p < 0.05 for both PEG and tracheostomy preferences). Patients scoring higher on the measure were also more likely to use tracheostomy in the follow-up period (p = 0.001). Attachment to life was lower among patients with longer duration of disease and lower at baseline among patients who died in the follow-up period (p < 0.01).
Physician communication to patients.
Our educational effort was uniform so that all patients received the same standardized educational session before the baseline interview. We could not control what physicians may have told patients and how this may have influenced patient preferences. In the accompanying survey of patient physicians, 29% said they “routinely” recommend tracheostomy. However, as stated earlier, we were unable to link physician recommendations to particular patients. Also, no information was available on whether patients switched physicians in the course of follow-up.
Discussion.
Deciding whether to have a tracheostomy is a central issue for ALS patients. Most studies have shown that the decision to be intubated or undergo tracheostomy in ALS is, for the most part, unplanned. In one series, over 90% of patients were intubated or had a tracheostomy in an emergency room6; in another series, less than 20% were said to have chosen mechanical ventilation in advance.10
In our prospective study, ALS patients in a tertiary care clinic had well-defined preferences for tracheostomy and PEG placement, and these preferences were related to actual use later. Patients in favor of the interventions were more likely to use them in the course of follow-up than were patients opposed.
Between 6% and 12% of this ALS cohort were committed to tracheostomy; they were certain or nearly certain that they would have tracheostomies, or stated that they wished to live as long as possible with tracheostomy. About a third of patients similarly stated they were in favor of PEG placement or nasal ventilation (BiPAP).
These preferences were elicited at a single point in time; we do not have measures of stability or change in such preferences. This is a limitation in our study and an area for future research. Evidence collected in this study suggests that preferences do change. For example, we found that patients with longer duration of disease were less likely to favor use of the technologies. Also, changing disease status may lead patients to change preferences. We found that patients with more rapid progression, as indicated by declining FVC values, were more likely to favor PEG (although preferences were not strongly related to disease status in cross-sectional analyses).
Preferences for life-extending or ameliorative technologies were also related to a patient’s attachment to life. Patients who stated that they had a strong will to live, had something to look forward to each day, continued to look to the future, etc., were more likely to favor tracheostomy and PEG placement. However, attachment to life may also change with disease progression. We found lower attachment to life among patients who later died or who had been living with ALS for a longer time.
Depression and hopelessness, by contrast, were unrelated to patient preferences. This finding should be considered in light of a study showing that hopelessness in ALS patients was related to more favorable attitudes toward assisted suicide.11 The difference may be due to our exclusion of patients with severe psychiatric disorders. Tracheostomy and PEG placement also differ from assisted suicide in that rejection of the interventions may reflect patient judgments about the quality of life impact of the technologies.
It is likely that other factors, unmeasured here, may also affect preferences. One set of factors involves physician recommendations, changes in medical care, or insurance coverage. Nearly a third of the physicians to these patients reported that they recommend tracheostomy at some point in the course of disease. This may affect patient preferences, even in the presence of the standardized clinical education effort used here. Also, we did not directly measure patients’ judgments about the quality of life impact of the interventions. For example, in the case of PEG, tolerance for time dedicated to eating (or discomfort from choking) may be an important determinant of early or later use of the technology. One patient, for example, reported that he accepted PEG placement because all his time revolved around eating, an intolerable quality of life infringement. Patients with a greater tolerance for such behavior or a different lifestyle might delay PEG under the same circumstances.
We found that ALS patients could articulate preferences for tracheostomy and PEG and that patient choices over the course of follow-up followed such preferences. We cannot say that this outcome was the result of our clinical education effort (that would require a clinical trial designed to investigate this explicitly). However, the common education effort, administered before the research interview, at least ensured that all patients had similar information to draw upon in formulating their preferences. Because our study suggests that patient preferences may change over the course of disease progression, a continued clinical educational effort is probably the best strategy.
These findings should be interpreted in light of our study design. We studied a single patient population, and all patients came from a single ALS center, where medical care is combined with a program of clinical education for patients and families. Clinical education efforts are likely to vary by site, and the relationship between patient preferences and clinical education in this population remains to be investigated. The opportunity for patients to realize their preferences in care decisions may also vary across sites. The determinants of such variation need further examination.
Acknowledgments
Acknowledgment
The authors acknowledge the contribution of the study Advisory Board (Drs. Ellen Katz, Landry Slade, Keith Burkum, Daniel Williams, and Karen Marder); Ms. Christine Weber, who conducted interviews; Mrs. Frieda Leopold, who volunteered many hours of counseling and support to patients and families; Dr. Judith Rabkin, for helpful suggestions in review of the paper; and patients and families receiving care at the Eleanor and Lou Gehrig MDA/ALS Center.
Footnotes
-
See also page 248
-
Research supported by the Project on Death in America and Fetzer Institute.
- Received August 5, 1998.
- Accepted March 22, 1999.
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
