Palliative Care Update for Primary Care Physicians
Authors: William J. Kennedy, DO, Fellow, Palliative Care Services, Mount Carmel Health, Columbus, OH; Philip H. Santa-Emma, MD, Medical Director, Palliative Care Services, Mount Carmel Health, Columbus, OH; and Robert M. Taylor, MD, Medical Director, Palliative Care Services, Mount Carmel Health, Columbus, OH.
Peer Reviewer: Lynn McDonald, MD, Medical Director, Hospice of Kankakee Valley, Bourbonnais, IL.
For all the triumphs of modern medicine, helping patients die well has not been one of them. In the last century, the median age of death has risen steadily to almost 80 years in some countries. The causes of death have shifted from infectious diseases to chronic illnesses (e.g., heart disease, cancer and stroke).
The nature of death itself has shifted along with these changes. What was once an acute or unexpected event, now is often a prolonged process, with progressive debility and increased suffering.1 Research has shown that many inpatients die in pain and at great financial and emotional cost to their families.2,3 Similarly, studies show that pain is under treated across the spectrum of clinical care, including long term care settings.4 The development of palliative medicine and its inpatient components, the palliative care service, and acute palliative care unit, are seen as primary solutions to these deficiencies in end-of-life (EOL) care. The word palliative, originating from the Latin word pallium, means to cover or to cloak. In its most general form, it relieves suffering in all arenas of care. Palliative medicine, in contrast to hospice, is not limited by prognostic estimates.
A palliative care service offers specialized consultation throughout the hospital, much as a cardiology or pulmonology service would function. It can range from a doctor and a nurse working in tandem to a more comprehensive team including a social worker, chaplain, and pharmacist. In more developed settings, a dedicated palliative care unit may be established. Such units, similar to other specialty intensive care units, focus on the acute palliative needs of patients. For example, a recently diagnosed patient with inoperable pancreatic cancer and severe abdominal pain would be a suitable candidate for palliative unit management. In such a setting, the patient could receive expert pain management along with team involvement to address the multitude of psychological, spiritual, and existential issues likely present.
Ideally, such care would be provided at time of diagnosis, with expert symptom control and psychosocial support complementing curative or disease modifying therapy. However, in a more practical vein, reasonable goals for the coming decade include the establishment of palliative care programs or services in every hospital, along with the continued spread of palliative practice standards in primary and specialty care. As Americans continue to live longer with chronic illness, primary care physicians increasingly will be required to identify suitable patients and to provide basic palliative care. In many ways, primary care physicians already provide the type of holistic care that is the hallmark of good palliative care. A key component in translating this patient centered approach to EOL care is improved education about symptom management, prognostication, and palliative specialist consultation.
Primary care physicians typically are interested in, and strive to be competent at, taking care of dying patients.5 This article identifies barriers to providing palliative care and, when possible, provides specific strategies for overcoming clinical and administrative problems. Specifically, the treatment of pain and dyspnea will be addressed with patient examples. Prognostication, EOL discussions, and the use of hospice criteria—all frequent areas of misunderstanding—will be examined from an evidence-based perspective. Finally, billing for palliative care will be addressed, detailing the under utilitzed option of billing for time spent with patients for counseling and education.
—The Editor
The Modern Death
Seventy five percent of modern deaths in the United States occur in institutional settings, such as hospitals or extended care facilities.6 In 1995, the landmark SUPPORT study of more than 4000 hospitalized dying patients found that 50% experienced moderate to severe unrelieved pain in the final days of their lives.2 Resuscitation status rarely was addressed until near death. Even when previously established, the physician often did not know the patient's wishes regarding EOL care. Many patients also spent their final days in the intensive care unit. Furthermore, the study intervention—the use of specially trained nurse clinicians to facilitate communication between patients and attending physicians—failed to produce any differences in patient care. One important response to this failure was the recognition of the need to develop structured programs in hospitals to care for the dying patient.
Palliative specialty consultation services and acute palliative care units assist attending physicians in all aspects of EOL care. Whether working on a dedicated unit or as a consult service, expert symptom management is paramount. Complementing this emphasis on symptom control is a team approach in assessing and managing the psychological and spiritual needs of the patient. A typical palliative care team may include a nurse clinician, a pharmacist, a chaplain, and a social worker. This interdisciplinary approach works best in dealing with the often complex needs of dying patients and their families. As the chaplain deals with issues of spiritual distress, the social worker can provide psychosocial support and help navigate insurance and discharge planning obstacles. The nurse clinician and palliative care physician, in addition to managing complex symptoms, can provide sensitive, but objective, advice concerning prognosis and goals of care.7,8
A key issue is understanding the relationship between disease directed therapy and palliative therapy. Disease-directed therapy (also sometimes called disease-modifying therapy or cure-oriented therapy) is the kind of medical therapy that physicians routinely provide to most of their patients. Palliative therapy, in contrast, is directed toward managing symptoms and other correlates, such as psychological and spiritual suffering, caused by the disease or its treatment. When a patient's disease is responsive to disease-directed therapy, and when the burdens of that therapy are not excessive, such therapy is not only appropriate, it constitutes the standard of care. As noted previously, palliative therapy ideally should be a component of the treatment of any serious illness from the time of diagnosis. However, if a patient's disease progresses to the point that it is no longer responsive to disease directed therapy, or if such therapy is perceived by the patient to be excessively burdensome, it is appropriate to transition to an exclusively palliative plan of care. Palliative care may, therefore, be appropriate for any patient with serious, life-limiting illness, either as a complement to disease-directed therapy or as the focus of care.
An acute palliative care unit may range from a few medical beds to a specialized area of the hospital. For patients with life limiting illness, care emphasizes symptom control and other issues associated with active, progressive disease. After the SUPPORT study, institutions across the country began to develop acute palliative care programs. In one significant effort, six model palliative care programs were selected by the Center to Advance Palliative Care (www.capc.org), with funding from the Robert Wood Johnson Foundation, to serve as Palliative Care Leadership Centers, and to assist other hospitals and hospices strengthen or establish their own programs. Palliative Care Leadership Centers represent a broad range of practice settings, from a community-based hospice to major academic medical centers.
The Mount Carmel Health System in central Ohio was chosen as a Palliative Care Leadership Center because of its extensive experience providing inpatient palliative care. The hospital system consists of three hospitals with approximately 1000 beds, a medical residency teaching program and a community-based hospice. In 1995, when it had a small number of inpatient hospice beds, its own internal review of EOL care found significant shortcomings, including hospice involvement in only 9% of its inpatient deaths. The Mount Carmel response included establishing a palliative consultation service along with designated inpatient palliative units in 1997. In its current form, the system maintains 17 total beds at three hospitals. It emphasizes creating a more home-like, less institutional environment. There is 24- hour visitation, with a focus on respecting patient and family privacy. Each unit also is equipped with a small kitchen and a family room.
To monitor its effectiveness, a database was created. It now has tracked the demographics and clinical outcomes of more than 7000 patients. This database is one of the largest series of inpatient palliative care patients in the United States. Between 1997 and 2004, involvement of the palliative care service in the care of patients who died in the hospital increased from 9% to 45%.Within four years of its implementation, the palliative care service was rated the most valuable patient service by attending medical staff.9 Since the publication of SUPPORT, multiple studies have confirmed the effectiveness of Mount Carmel style inpatient programs.10-12
The Acute Palliative Care Unit (APCU) and Primary Care Practice
The clinical and personal characteristics required to provide excellent primary care overlap considerably with the requirements for providing effective palliative care. Both specialties focus on treating the entire patient, including the emotional, spiritual, and familial realms.13 Nevertheless, most physicians report feeling unprepared to care for the dying,14-16 which reflects, in part, a lack of formal education and exposure.17 However, despite concerns about their ability to provide palliative care themselves, fear that palliative care physicians will assume primary patient care responsibilities has been cited as an inhibiting factor in accessing palliative care consultation services. In reality, primary care physicians can learn to provide competent palliative care while also utilizing a palliative care service for difficult cases.18,19 Palliative care involvement should be seen as analogous to other consultation services, providing specialist care for specific components of the patient's illness.
Also, the interdisciplinary approach of palliative care teams distributes responsibility for the intense psychological needs of dying patients and their families, which have been shown to be a principal source of primary care physician burnout.20 APCU team involvement improves patients' insights into their diagnosis and prognosis.21,22 This support may reduce chances of patients becoming "frequent flyers" with recurrent exacerbations of essentially end-stage illnesses, especially those with non cancer disease. For those primary care physicians who utilize hospitalists for inpatient care, the palliative consult service approach facilitates post-discharge care planning and enhances communication between hospital and primary care physician as the patient returns to the outpatient setting. With as many as half of all primary care visits involving psychosomatic complaints,23-24 these interventions may be of considerable importance. Overall, APCU/PC improves symptom control, patient insight, and cost effectiveness. In the outpatient setting, palliative care has been shown to reduce urgent care and primary care visits without associated increases in specialist utilization.25 Given these results, it is not so surprising that family physicians and internists give palliative care and hospice very favorable ratings in terms of their involvement in patient care.26
Primary Care as Outpatient Palliative Care
(All patient names in this article are fictitious.)
Pain Case Study. Mr. Reynolds was a 63-year-old male with a history of chronic renal insufficiency with a serum creatinine level of 3.3 mg/dL, an estimated creatinine clearance of 24, diabetes, and a recent diagnosis of inoperable pancreatic cancer. He has been taking two Vicodin (5/500) tablets every 4 hours as needed for non radiating, stabbing-type upper quadrant abdominal pain. He reports the pain has been increasing, ranging from a 3/10 30 minutes after taking his Vicodin, to an 8/10 immediately before his next dose. Due to the severity of his pain, he took the maximum prescribed dose (i.e., 12 tablets per each day), for the past week. The patient also has had burning sensations in his legs secondary to his diabetes, for which the Vicodin has not been helpful. He is interested in achieving better control for his abdominal pain along with initial treatment for his lower extremity neuropathic pain. How should this patient be approached?
In 1990, the World Health Organization (WHO) published its guidelines for treating cancer pain.27 The WHO algorithm (Figure 1) advocates using a step-wise approach, with stronger opioids combined with adjuvants for increasing pain. A 10-year follow-up study of the guidelines involving 2118 patients studied for a total of 140,478 treatment days, showed highly significant pain reduction in the first week of treatment; good to satisfactory pain relief, per patient report, was achieved in 88% of patients during the whole treatment period.28 When compared with the results of the SUPPORT study—where half of the patients had moderate to severe pain until their deaths—the results were especially convincing.
Beyond the lack of effective analgesia, Mr. Reynolds has other problems with his current regimen. The most common oral formulations of hydrocodone (e.g., Vicodin and Lortab) contain 500 mg of acetaminophen per tablet in addition to the opioid. Patients without liver disease should not take more than 4 grams of acetaminophen per day. Therefore, Mr. Reynolds is taking a potentially toxic dose of acetaminophen. Also, in cases of uncontrolled pain, PRN (as needed) dosing fails to achieve steady state levels of medication, yielding erratic pain control.
As pain is subjective by definition, all pain control is empiric. The right dose and medication regimen are defined by what works for the patient. Said another way, there may be several successful medication regimens that will achieve adequate analgesia in a given patient. However, the best approaches all adhere to the same principles. Common prescribing errors include failing to consider end-organ function (e.g., morphine use in cases of renal impairment) and the failure to establish around-the-clock dosing. Of course, the overall clinical picture must be considered. In certain situations, pain control or symptom management may be a higher priority than preserving organ function. For example, the use of intravenous ketorolac may be the treatment of choice in cases of severe, end stage bone pain, despite the deleterious effects on renal function.
In the case of Mr. Reynolds, his chronic renal insufficiency is a relative contraindication to the use of morphine or non steroidal anti-inflammatory drugs (NSAIDs). Morphine's active metabolite (3-6 gluconoride) accumulates with decreased renal function, which may cause neuro excitation and even seizure activity. NSAIDs, which are helpful particularly with bone pain, cause decreased prostaglandin mediated dilation of the renal artery, decreasing renal perfusion and, therefore, may cause further renal failure. Given his high functional status, preserving renal function is an important consideration of this patient.
When pain is continuous, pain tolerance actually is decreased with only PRN-type dosing for analgesia. As pain is allowed to increase, the patient often requires greater doses to control the exacerbation. Hence, a cycle of pain exacerbation and relief is established, creating a state of anticipatory pain in the patient. A more effective regimen, often with less analgesic medication per 24 hours, involves using an around-the-clock dose as a baseline. Breakthrough pain is treated with PRN dosing, which also serves as a useful reference to calculate daily medication needs, permitting safe and effective titration of the scheduled dosing.
Regardless of regimen, patients fare best when educated about the nature of illness-related pain and the effective use of analgesic medicines. Particularly older patients, unwarranted concerns about potential addiction or tolerance may result in underreporting pain. Dispelling these concerns requires eliciting their perceptions about pain and opioid use. Increasing opioid usage most often represents disease progression, not medicine tolerance. A simple reassurance of that fact may be all that is necessary, although more extensive discussion and reassurance sometimes may be required. In the case example, the best practice is to clarify the nature of the patient's pain and also explain the rationale for the treatment plan and why it is believed it will be effective.
In accordance with the WHO guidelines, start with an adjunctive medicine (e.g., gabapentin) for this patient's diabetes-related neuropathic pain. Gabapentin is excellent first line therapy and avoids the side effects associated with less effective medicines (e.g., acetaminophen and NSAIDs). Opioids also are effective therapy for neuropathic pain. Although one could make an argument against adding a second drug —gabapentin in this case— and only increasing the opioid medication, the risks of adding gabapentin are quite low, and there may be a benefit in providing better pain control with a lower opioid dose. Because Mr. Reynolds already has failed a combination of hydrocodone and acetaminophen, continuing up the WHO ladder, we will schedule a strong opioid around the clock, at a higher equivalent dose than his current usage. In the United States, hydrocodone is only available as a combined formulation with acetaminophen. Therefore, we will use a similar synthetic morphine derivative (e.g., oxycodone), which is available without acetaminophen. As with hydrocodone, it does not have the same neuroexcitatory side effects associated with morphine in renal failure.
Opioid Conversion. To convert from one opioid to another, start by calculating the total 24-hour usage. Next, convert the 24-hour total to an equivalent amount of the new opioid.
Among the most commonly used oral opiods for cancer pain, morphine, oxycodone, and hydrocodone have different references and provide somewhat different equivalencies. However, more recent publications have indicated that oral oxycodone is approximately 50% more potent than oral morphine (e.g., 10 mg of oral oxycodone is roughly equivalent to 15 mg of oral morphine).29 Although there is much less empirical evidence as to the relative potency of hydrocodone, a 1:1 ratio between oxycodone and hydrocodone commonly is used in clinical practice without any problem.
Therefore, Mr. Reynolds equivalent oxycodone dose also can be considered 60 mg. When converting between different opioids, there is only partial cross tolerance. Therefore, an equivalent dose of a new opioid generally will be stronger than the previous opioid. To account for that effect, when pain is controlled, convert to a dose that equates to 50-75% of the original opioid. However, in Mr. Reynolds' case, the patient does not have controlled pain. Generally, opioids should be increased by 33-50%for uncontrolled pain. We account for these factors—partial cross tolerance and increased pain—by converting at 100% equivalence. In this way, the lower dosing involved with partial cross tolerance is matched by the increased opioid need for worsening pain.
Mr. Reynolds has used two Vicodin tablets every four hours, a total of 12 tablets per day or 60 mg of hydrocodone per day. As explained above, we, therefore, will convert him to a total daily dose of 60 mg oxycodone. We have two options in choosing his oxycodone regimen. We can start him on a scheduled dose of short-acting oxycodone, called OxyIR (which, based on its half life, is dosed every 4 hrs) at a dose of 5 mg every 4 hours for a total of 60 mg/day. However, a second option would be to prescribe a long-acting form, OxyContin, divided into two or three doses per day (e.g., 30 mg every 12 hours). When pain control is stabilized, the longer acting OxyContin often is preferred due to administration convenience for the patient. In either case, a breakthrough dose of approximately 5-15% of the 24-hour dose is prescribed every one hour PRN. In our example, the breakthrough dose would be 5-15% of 60 mg, or 3 to 9 mg. Because OxyIR is available as a 5-mg tablet, the dose usually would be rounded to 5 mg.
Whenever any opioid is prescribed, a bowel regimen must be started to avoid constipation. In contrast to other opioid side effects such as sedation and nausea, which usually resolve within 24-48 hours, a patient never becomes tolerant to the constipating effects of opioids. A recommended approach involves using a stool softener (e.g., colace) along with an activating laxative, (e.g., senna) in a combined formulation. Senokot-S combines these agents and can be given as either by mouth or as a suppository per rectum. A typical beginning oral regimen is one capsule BID. If necessary the dose may be doubled to two capsules BID. For severe constipation, stronger laxatives, lactulose, suppositories, and/or enemas may be required. Table 1 summarizes the above noted approach.
Legal Issues in Pain Management. Controlled substance regulations are believed to inhibit physicians from prescribing opioids for patients with cancer pain.30 A complete discussion about legal and regulatory issues in pain management is beyond the scope of this article. However, several important points can be made. First, specific guidelines exist for the appropriate use of controlled substances for pain. Table 2 defines the most recent treatment recommendations from the Federation of State Medical Boards.31 In general, treatment of nonmalignant pain is scrutinized more closely than treatment for malignant pain. However, in either case, the principles of appropriate dosing documentation apply. In many ways, standard of care pain management simply is practicing good medicine: appropriate evaluation of the patient, development of a treatment plan with periodic review, keeping accurate medical records, and referring the patient for consultation as necessary.
Although sanctions for inappropriate and excessive prescribing of pain medicines has been well publicized, under treating pain is now a legal liability as well. In 2001, a California jury found an internist guilty of elder abuse for failing to effectively treat a terminally ill patient's pain.32 Therefore, now more than ever, learning appropriate pain management documentation and techniques is essential for the primary care physician.
The use of intensols (i.e., concentrated liquid preparations given sublingually) is an important component in managing patients receiving palliative care. As a patient approaches death, increasing lethargy and dysphagia present challenges in medication delivery. Intravenous administration often is not feasible or desirable in the extended care facility and home environments. Commercially available intensols (Table 3) include morphine5 (20mg/mL), oxycodone (20mg/mL), haloperidol (2mg/mL), lorazepam (2mg/mL), and diazepam (5 mg/mL). Hence, the use of intensols often will ameliorate the most common symptoms associated with advanced disease —pain, dyspnea, nausea, and agitation—without IV access or the ability to swallow.
Dyspnea. Dyspnea is a common, distressing symptom in dying patients. In contrast to pain, dyspnea tends to increase in prevalence and severity as death approaches.33 Causes of dyspnea are multiple. The average number of causes in one series of 100 patients with cancer and dyspnea was five. Fewer than half of those patients were hypoxic, with an oxygen saturation level below 90%.34 Simple tests (e.g., pulse oximetry, blood count, and chest x-ray) along with physical examination usually are sufficient to identify reversible causes.35 The use of laboratory tests and diagnostic studies must be considered in the frame of reference of the patient's goals. It may not be appropriate to monitor laboratory test results in a patient who is imminently dying. It may be appropriate to monitor chemistry results in a patient with congestive heart failure who is receiving palliative care. One of the greatest challenges clinicians face, intellectually and emotionally, is determining when it is appropriate to treat a patient empirically.
Oxygen is the treatment of choice if the patient is symptomatic exclusively from hypoxemia. However, many EOL patients have multiple causes of dyspnea, and oxygen will not be the only treatment needed. In most cases, the exact role of oxygenis unclear. As noted, dyspnea is not well correlated with hypoxemia in cancer patients. Oxygen's effect in many patients may be from its direct sensory stimulation. Studies have shown that cool air blowing on the face can reduce dyspnea.36 This phenomena is thought to be regulated by the trigeminal nerve, such that stimulation may have a central inhibitory effect on dyspnea.
The pharmacologic treatment of choice for dyspnea related to chronic, severe illness is opioids.35 Low doses (e.g., 2.5-5 mg of oral morphine every 4 hours) may be all that is needed in mild dyspnea. In patients with chronic obstructive pulmonary disease, small doses with appropriate titration can be used without significant risk of respiratory depression. A typical dosage for an opioid-naive patient would be to start with a scheduled dose of morphine 2.5 mg PO every 4 hours, with 2.5 mg PO every one hour PRN for dyspnea. The dose may be increased by 50-100% every 24 hours PRN (e.g., to 5 mg, then, to 7.5 or 10 mg). When dyspnea is not responsive to conservative dosing or when dosage needs change rapidly, as with an actively dying patient, palliative care consultation can be helpful.
Anxiolytics can be helpful in patients when anxiety is a primary component of a patient's symptomatology, but are of limitedvalue otherwise.37,38 Bronchodilators can cause anxiety via sympathetic stimulation; therefore, if their use does not appear to relieve dyspnea, consider discontinuing them.
Beyond symptomatic medicines, consider reversible, anatomic causes. Palliative chemotherapy, radiation, and resection all can relieve symptoms without further interventions but may involve significant risks in very ill patients. Airway obstructions can be stented; pleural effusions can be drained. However, the clinician must attempt to determine—with the help of specialist input if necessary—when interventional techniques are worthwhile and appropriate for each patient. As a patient's illness progresses, the risk-benefit ratio gradually will shift toward strict symptom management.
This fact is particularly true in the acute dying process, where the changes in breathing patterns, including the "death rattle," can be particularly troubling for families. Preventive measures include minimizing intravenous fluids, which tend to accumulate in the airways in actively dying patients, along with positioning the head in a more upright position. Treatment options start with deciding whether it is possible to expectorate the secretions. If so, recommendations include encouraging oral hydration, adding guafenes in liquid (10mL PO q4), and nebulized saline (0.9% 3mL inhaled q4-6). If the patient is unresponsive or unable to expectorate secretions, anticholinergic agents (e.g., Hyoscyamine 0.125-0.25-mg tablets can be crushed, mixed with a small amount of water, and given sublingually every 4-6 hours PRN.) or scopolamine patch (e.g., 1.5 mg transdermal every 72 hours) are often effective.
Anticipating the Uncertain: Referral and Prognostication
Barriers to Referral. The reasons for lack of palliative care and hospice referral are complex. A 2004 systematic literature review of barriers to referral included physician and nursing issues, nature of disease, ethnicity and cultural issues, patient age, and lack of access to services.39 Reasons for physician non referral included inadequate knowledge of hospice policies and practices, inadequate knowledge of death and dying issues, negative attitudes toward hospice care, and concern about loss of control of patient care after referral.40-43
Prognosis. Of these issues, prognostication may be the most significant factor in causing late hospice and palliative referral. Physicians report finding it challenging to understand and respond to the information needs of their patients, which can lead to unintended suffering and regret.44 In one sample of more than 2300 cancer patients, 87% wanted to know their prognosis.45Yet, doctors in many specialties find it difficult to prognosticate and have a consistent bias toward over-optimism. In a large, prospective cohort study, 80% of physician prognostications were incorrect, with an overall overestimated survival by a factor of five.46 Interestingly, the stronger the physician-patient relationship, the worse the prognostic accuracy.
When patients do not have accurate prognostic information their ability to address a host of essential EOL issues is limited. Such issues encompass a range of personal matters, including establishing care goals and advance directives, attending to legal and financial matters, and focusing on closure and life legacy issues. In one series, psychiatric distress was highest in terminally ill patients who were not told the truth about their condition and prognosis.47 Partially because of these difficulties, patients typically receive less than one month of care in hospice,48 although shorter length of hospice care is correlated highly with lower patient family satisfaction.49
Patients and physicians often have different expectations about illness prediction. In a large sample of attitudes and beliefs regarding terminally ill patients, internists reported feeling both stressed and poorly prepared to prognosticate. They were concerned that patients may have unrealistic expectations and judgments regarding the accuracy of their prognosis.50 In fact, it is most reasonable to give prognoses in ranges (e.g., hours to days, days to weeks, weeks to months, or months to years). Even then, it is wise to emphasize that any prognosis is statistical and that any particular patient may live longer or shorter. This approach minimizes the risk of error and provides the patient and family with sufficient information to make reasonable life decisions. Indeed, family members have reported feeling less stress when provided with probabilities and ranges rather than definite estimates of survival.44
In non cancer patients, the use of specific predictions may be especially problematic. Follow-up analysis of the SUPPORT study data found that clinical prediction criteria were not effective in identifying six-month survival in patients with advanced lung, heart, or liver disease.51 In highly variable diseases such as amyotrophic lateral sclerosis (ALS), strict hospice admission criteria actually may exclude the majority of eligible patients from appropriate hospice referral.52
Primary care doctors, with their long term, highly invested relationships with patients, are at the center of this issue. With the majority of U.S. hospice patients now admitted with non cancer diagnoses, appropriate and reasonable efforts at prognostication of chronic illnesses becomes increasingly important. Primary care physicians often are in the best position to foresee the inexorable but erratic decline in patients with such conditions. A key to this process is reframing the way physicians view such elements as performance status and illness progression in their patients.
New Perspectives: Rethinking Hospice Criteria. Performance status has been shown to be the best predictor of survival in cancer patients. Among the various performance scales, the Karnofsky scale (Table 4) is the oldest and best known. However, the newer Palliative Performance Scale (PPS), developed by Victoria Hospice Society (Table 5), also is well-validated and now is used more widely in EOL functional assessment. Having been developed before the widespread use of hospices, the primary disadvantage of the Karnosky scale is that it includes the "necessity for hospitalization" as one of the factors to be considered. In contrast, the PPS avoids that anachronism and includes only easily evaluated clinical factors. Furthermore, the PPS specifically was developed to correlate as closely as possible with the Karnofsky scale. For patients with cancer, a Karnofsky score orPPS of 50% (i.e., requires considerable assistance with ADL) or less suggests a life expectancy of two months or less.53 Likewise, in non cancer disease, frequent hospitalizations and being symptomatic despite maximal disease-directed medical therapy are superior predictors of poor prognosis.8
Medicare has addressed the problematic nature of eligibility criteria in revising its hospice guidelines. Whereas, objective hospice criteria (Table 6) originally were emphasized, more recent guidelines focus on physician judgment. Current guidelines acknowledge both the difficulty of patient prognosis and the lack of risk in certifying a patient as terminally ill.54 Given these changes, one should interpret hospice criteria as suggestions for inclusion rather than for exclusion. In a busy primary care practice, these criteria may serve as reminders in identifying patients who may be better served with hospice care.
Dual Prognostication. The dual prognostication model was developed in response to the problems inherent in predicting outcomes in chronically ill, nonc ancer patients. The model promotes patient involvement in both palliative and curative therapy as early as possible in their illnesses.55 Figure 2 shows different illness trajectories over time. Care combines disease-directed treatment along with comprehensive palliative measures. Such care would include symptomatic relief, resuscitation status clarification, and living will completion. As patients begin to fail medically, the transition to comfort-focused care would be both familiar and gradual. The backbone of such an approach involves communication skills regarding difficult EOL topics and expert symptom management.
Research has defined specific communication styles shown to work well in EOL discussions. In Ambuel's living well interview, an open-ended approach is encouraged.56 His recommendations come from observations that EOL conversations often are framed mistakenly as a choice between treatment and withdrawal. Instead, Ambuel recommends using an exploratory approach (Table 7).57 Firstly, develop a broad understanding of the patient's hopes and goals. In this way, the patient and family define the objectives; the physician helps define probable outcomes and what treatments likely will be consistent with the patient's goals.
All discussions about EOL care and goals should start with assessing the patient's understanding of his or her illness and prognosis. A typical discussion includes four general factors: 1) pain and non pain symptom assessment and management issues; 2) assistance with difficult decisions (e.g., feeding tubes, ventilator withdrawal); 3) addressing most appropriate care setting for the patient; and 4) providing psychological support.56 A recommended approach involves starting with more practical and immediate issues such as symptom management. Once these topics are addressed, the conversation often segues naturally into more emotion-laden issues (e.g., resuscitation status and EOL care goals). This open-ended approach acknowledges both the patient's individual experience of his or her illness and helps prevent the doctor from making erroneous assumptions about the patient's understanding or wishes.
Discussing hospice care can be especially difficult, given its emotional connotations with many patients. One suggested approach involves presenting the benefits—such as expert symptom management and home assistance—before naming the provider as hospice. Table 8 outlines a simple step-wise approach in discussing hospice care, which largely parallels the above noted approach in discussing EOL care.58 A key component involves exploring the patient's preconceptions about what hospice care involves. These assumptions, which may involve previous experiences with family or friends, as well as more general issues related to terminal diagnosis, can be the source of significant anxiety or resistance regarding hospice referral. Once the patient's assumptions and concerns have been addressed, explaining the benefits of hospice referral can be tailored to the patient's needs and concerns.
Billing for Palliative Care. The same general principles used for billing for non palliative patients apply to billing for palliative patients, with a few minor caveats.59 For primary care doctors, palliative management could involve symptom management as well as patient counseling, education, and coordination of care. When the visit (either inpatient or outpatient) focuses on medical management, including symptom management, the usual E&M codes ordinarily should be used and will result in a fair and reasonable reimbursement for the physician's services. However, when the majority of time spent during a visit is devoted to education and/or counseling of the patient—either inpatient or outpatient —or to coordinating inpatient care, it is more appropriate to bill for time; this approach will result in fairer reimbursement for the physician's work. It is important to note that billing for time is an option whenever the physician devotes the majority of his or her time to such efforts, not only when these issues involve EOL discussion.
There are no specific billing codes for addressing such EOL issues as establishing code status or delineating advanced care directives. Because these efforts typically are relatively time consuming and may be the focus of that visit, billing for the medical aspects of the encounter would result in unfair compensation. To be compensated for such work, the physician usually should code for time. When billing for time, the guidelines are slightly different for inpatients and outpatients (Table 9).
In the outpatient setting, total time is defined only by actual face-to-face time spent with the patient. If more than 50% of this face-to-face time is spent counseling or educating the patient, then, time can be used as the sole billing criterion.
For example, a primary care physician spends 40 minutes with a patient after recent imaging and biopsy studies revealed widely metastatic prostate cancer. During the visit, approximately 15 minutes are spent reviewing medications, discussing symptomatic issues, and performing a physical examination. The final 25 minutes of the visit involve the physician answering questions about diagnosis and prognosis as well as providing counseling about EOL issues. The physician, then, would code the visit as a 99215, based on total time spent, with more than 50% of the visit consisting of counseling and information giving. The physician specifically would document in the note that a total of 40 minutes was spent with the patient and that more than 50% (i.e, 25 minutes) was spent in counseling and information giving (e.g. "total time spent was 40 minutes, more than 50% spent counseling and giving information."). In addition, the physician should summarize topics addressed (e.g., prognosis, hospice care, and resuscitation status) along with noting that these issues needed to be addressed because of the recent diagnosis of metastatic prostate cancer. New patient visits and consults in the outpatient setting also can be coded based on time using the same guidelines.
In an inpatient setting, physicians should bill for time if more than half of the total time spent managing the patient is devoted to counseling, information giving, and/or coordination of care. For example, suppose the above patient was seen in an inpatient setting, and a total of 70 minutes was spent managing the patient, divided as follows: reviewing test results (10 minutes), discussing symptoms and examining the patient (15 minutes), adjusting medications (5) minutes), discussing the case with the oncology consultant (5 minutes), answering questions about diagnosis and prognosis, as well as providing counseling about EOL issues (30 minutes), and discussing the case with the social worker to initiate a hospice referral (5 minutes). In this case, 40 minutes (i.e., the sum of the last three elements) were spent in counseling, giving information, and coordination of care. Thus, the physician could bill 99233 and 99256 for that day's work. As in the previous example, the physician also should document the time spent (e.g. "total time spent was 70 minutes, more than 50% spent counseling, giving information, and coordination of care") and summarize topics addressed and the reason that these issues needed to be addressed. New patient admissions and consultations in the inpatient setting also can be coded based on time using the same guidelines.
Conclusion
During the next 25 years, the population of persons age 65 years and older will double. As medical technology advances producing further increases in longevity, the number of patients living with chronic severe illness will continue to increase. The need for competent palliative care across specialties will grow proportionally with these demographic shifts. By the end of the 20th century, alarming results from studies such as the SUPPORT study found a health care system largely divorced from the needs of chronically ill and dying patients.2
Fortunately, sustained institutional efforts have developed effective models of care to address these shortcomings. The palliative care service—as a consult team and as a specialized unit—is a valid approach to meeting the needs of severely ill and dying patients. Increasingly, the concepts and techniques of palliative care will spread from specialty services to primary care practice. With the anticipated demographic changes, these developments will be essential to meeting the needs of the aging population.
Palliative care—like all specialty care—is a learned skill set that can be applied effectively by primary care physicians. Symptom management in palliative patients, such as with pain and dyspnea, involves a logical treatment algorithm, just as one would treat hypertension or dyspepsia. More complicated issues, such as disease prognostication, require a combination of knowledge of illness predictors with awareness of the potential for cognitive bias, especially in patients where there is a long-term invested relationship. Despite these challenges, primary care physicians are capable and well positioned to address the whole person needs of the aging population. Indeed, in the coming decades, they must play a central role if we are to ensure that the elderly traverse the final stages of their lives with dignity and a minimum of suffering.
References
1. Fischberg D, Meier D. Palliative care in hospitals. Clin Geriatr Med 2004;20:735-751.
2. The SUPPORT Principal Investigators: A controlled trial to improve care for seriously ill hospitalized patients: The Study to Understand Prognoses and Preferences for Outcomes and Risks for Treatments (SUPPORT). JAMA 1995;274:1591-1598.
3. Covinsky K, Goldman L, Cook F, et al. The impact of serious illness on patients' families. JAMA 1994;272:1839-1844. 4. Ferrell B. Pain evaluation and management in the nursing home. Ann Intern Med 1995;123:681-687.
5. Steinmetz D, Walsh M, Gabel L, et al. Family physicians involvement with dying patients and their families. Arch Family Med 1993;2:753-760.
6. Teno J, Clarridge B, Casey V, et al. Family perspectives on end-of life care at the at the last place of care. JAMA 2004;291:88-93.
7. Santa-Emma P, Roach R, Gill M, et al. Development and implementation of an inpatient acute palliative care service. J Palliat Med 2002;5:93-100.
8. Weissman D. Consultation in palliative medicine. Arch Intern Med 1997;157:733-737.
9. Physician Update. A Mount Carmel Health System Publication. 1999.
10. Hanks G, Robbins M, Sharp K, et al. The imPaCT study: A randomized controlled trial to evaluate a hospital palliative care team. Brit J Cancer 2002;87:733-739.
11. Edmonds P, Stuttaford J, Penny J, et al. Do hospital palliative care teams improve symptom control? Use of a modified STAS as an evaluation tool. Palliat Med 1998;12:345-351.
12. Hearn J, Higginson I. Do specialist palliative care teams improve outcomes for cancer patients? A systematic literature review. Palliat Med 1998;12:317-332.
13. Block S, Bernier G, LaVera C, et al. Incorporating palliative care into primary care education. J Gen Intern Med 1998;11:768-773.
14. Von Gunten C,Von Roenn J, Neely K, et al. Hospice and palliative care: Attitudes and practices of the physician faculty of an academic hospital. Am J Hospice Palliat Care 1995;12:38-42.
15. Steinmetz D,Walsh M, Gabel L, et al. Family physicians involvement with dying patients and their families. Arch Family Med 1993;2:753-760.
16. Schneiderman L. The family physician and end-of-life care. J Fam Pract 1997;45:259-262.
17. Oneschuk D, Moloughney B, Jones-McLean E, et al. The status of undergraduate palliative medicine education in Canada: A 2001 survey. J Palliative Care 2004;20:32-37.
18. Ramsay A. Care of cancer patients in a home-based hospice program: A comparison of oncologists and primacy care physicians. Fam Pract 1992;34:170-174.
19. Desmedt M, Michel H. Palliative home care: improving co-operation between the specialist team and the family doctor. Support Care Cancer 2002;10:343-348.
20. Winefield H, Anstey T. Job stress in general practice: Practitioner age, sex and attitudes as predictors. Fam Pract 1991;8:140-144.
21. Jack B, Hillier V,Williams A, et al. Hospital based palliative care teams improve the insights of cancer patients into their disease. Palliat Med 2004;18:46-52.
22. Ellershaw J, Peat S, Boys L. Assessing the effectiveness of a hospital palliative care team. Palliat Med 1995;9:145-152.
23. Kroenke K, Mangelsdorff D. Common symptoms in ambulatory care: incidence, evaluation, therapy, and outcome. Am J Med 1989;86:262-266.
24. Barsky A, Borus J. Somatization and medicalization in the era of managed care. JAMA 1995;274:1931-1934.
25. Rabow M, Dibble S, Pantilat S, et al. The Comprehensive Care Team: A controlled trial of outpatient palliative medicine consultation. Arch Intern Med 2004;164:83-91.
26. Ogle K, Mavis B,Wang T. Hospice and primary care physicians: Attitudes, knowledge, and barriers. Am J Hospice Palliat Care 2003;20:41-51.
27. WHO Expert Committee. Cancer pain relief and palliative care. WHO Technical Report Series 1990; 2-75.
28. Zech D, Grond S, Lynch J, et al. Validation of World Health Organization guidelines for cancer pain relief: A 10 year prospective study. Pain 1995;63:65-76.
29. Doyle D, Hanks G, Cherny N, et al. Oxford Textbook of Palliative Medicine, 3rd ed. Oxford: Oxford University Press; 2004.
30. Weissman D. Doctors, opioids, and the law: The effect of controlled substances regulations on cancer pain management. Semin Oncol 1993;20:53-58.
31. Federation of State Medical Boards of the United States, Inc. Model policy for the use of controlled substances for the treatment of pain. Available at: http://www.fsmb.org. Accessed January, 2005.
32. Crane M. Treating pain: Damned if you don't? Med Econ 2001; 66-68. 33. Higginson I, McCarthy M. Measuring symptoms in terminal cancer: are pain and dyspnea controlled? J Research Social Med 1989;82:264-267.
34. Dudgeon D, Lertzman M. Dyspnea in the advanced cancer patient. J Pain Symptom Manage 1998:162:212-219.
35. Thomas J, von Gunten C. Management of dyspnea. Support Oncol 2003;1:23-34.
36. Schwartzstein R, Lahive K, Pope A, et al. Cold facial stimulation reduces breathlessness induced in normal subjects. Am Rev Respir Dis 1987;36:58-61.
37. Man G, Hsu K, Sproule B. Effect of aprazolam on exercise and dyspnea in patients with chronic obstructive pulmonary disease. Chest 1986;90:832-836.
38. Woodcock A, Gross E, Geddes D. Drug treatment of breathlessness: Contrasting effects of diazepam and promethazine in pink puffers. BMJ 1981;283:343-346.
39. Ahmed N, Bestall J, Ahmedzai S, et al. Systematic review of the problems and issues of accessing specialist palliative care by patients, careers and health and social care professionals. Palliat Med 2004;18:525-542.
40. Bradley E, Cramer L, Bogardus S, et al. Physicians' ratings of their knowledge, attitudes, and end-of-life care practices. Acad Med 2002;77:305-311.
41. Bradley E, Fried T, Kasl S, et al. Referral of terminally ill patients for hospice: Frequency and correlates. J Palliat Care 2000;16:20-26.
42. Miller K, Miller M, Single N. Barriers to hospice care: Family physicians perceptions. Hospice J 1997;12:29-41.
43. Ogle K, Mavis B, Wang T. Hospice and primary care physicians: Attitudes, knowledge, and barriers. Am J Hospice Care 2003;20:41-51.
44. Morita T, Akechi T, Ikenaga Y, et al. Communication about the ending of anticancer treatment and transition to palliative care. AnnOncol 2004;15:1551-1557.
45. Jenkins V, Fallowfield L, Saul J. Information needs of patients with cancer: Results from a large study in UK cancer patients. Brit J Cancer 2001:84:48-51.
46. Christakis N, Lamont E. Extent and determinants of error in doctors' prognoses in terminally ill patients: Prospective cohort study. BMJ 2000;320:469-473.
47. Gerle B, Lunden G, Sandblom P. The patient with inoperable cancer from the psychiatric and social standpoints. Cancer 1960;13:1206-1217.
48. National Hospice and Palliative Care Organization. Information Central/Hospice Statistics. Available at: www.nhpco.org/files/public/Hospice_Facts_110104.pdf. Accessed January, 2005.
49. Stillman M, Syrjala K. Differences in physician access patterns to hospice care. J Pain Symptom Manage 1999;17:157-163.
50. Christakis N, Iwashyna T. Attitude and self-reported practice regarding prognostication in a national sample of internists. Arch InternMed 1998;158:2389-2395.
51. Fox E, Landrum-McNiff K, Zhong Z, et al. Evaluation of prognostic criteria for determining hospice eligibility in patients with advanced lung, heart, or liver disease. JAMA 1999;282:1638-1645.
52. McCluskey L, Houseman G. Medicare hospice referral criteria for patients with amyotrophic lateral sclerosis: A need for improvement. J Palliat Med 2004;7:47-53.
53. Reuben D, Mor V, Hiris J. Clinical symptoms and length of survival in patients with terminal cancer. Arch Intern Med 1988;148:1586-1591.
54. Centers for Medicare and Medicaid Services. Available at: www.cms.hhs.gov/providers/hospiceps. Accessed: January2005.
55. Lynn J, Harrel F, Cohn F, et al. Prognoses of seriously ill hospitalized patients on the days before death: Implications for patient care and public policy. New Horiz 1997;5:56-61.
56. EPERC: End of Life/ Palliative Education Resource Center. FastFacts and Concept #065: The Living Well Interview. Available at: www.eperc.mcw.edu/fastFact/ff_65.htm.
57. EPERC: End of Life/ Palliative Education Resource Center. FastFacts and Concept #042: The Living Well Interview. Available at: www.eperc.mcw.edu/fastFact/ff_42.htm.
58. EPERC: End of Life/ Palliative Education Resource Center. FastFacts and Concept #038: The Living Well Interview. Available at: www.eperc.mcw.edu/fastFact/ff_38.htm.
59. Von Gunten C, Ferris F, Kirschner C, et al. Coding and reimbursement mechanisms for physician services in hospice and palliative care. J Palliat Med 2000;3:157-164.
Authors: William J. Kennedy, DO, Fellow, Palliative Care Services, Mount Carmel Health, Columbus, OH; Philip H. Santa-Emma, MD, Medical Director, Palliative Care Services, Mount Carmel Health, Columbus, OH; and Robert M. Taylor, MD, Medical Director, Palliative Care Services, Mount Carmel Health, Columbus, OH. Peer Reviewer: Lynn McDonald, MD, Medical Director, Hospice of Kankakee Valley, Bourbonnais, IL.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content