Medical care issues for primary care physicians
David Yu, MD
Preview: The complex management issues related to spinal cord injury traditionally have been the purview of physical medicine and rehabilitation specialists. However, changes in the healthcare system now offer primary care physicians an expanded role in helping affected patients live a healthier and more functional life. With proper understanding of the mechanisms of spinal cord injury, primary care physicians can become important members of the medical management team. Dr Yu presents a comprehensive overview of medical care issues and common complications in spinal cord injury.
Spinal cord injury is a devastating event that has far-reaching functional and psychological effects on the survivor. Only through optimal medical and rehabilitative care can such patients regain a sense of control over their life and maximize their independence and quality of life.
Historically, rehabilitation specialists have had primary responsibility for both medical care and rehabilitation of patients with spinal cord injury. However, because of geographic constraints and trends in healthcare, primary care physicians have a growing role in the care of such patients.
Injury to the spinal cord affects organ systems both above and below the level of injury, creating a wide spectrum of medical issues. Furthermore, both the medical and the rehabilitative issues vary according to the chronicity of the injury. A detailed discussion of all the relevant medical issues is beyond the scope of this article. Instead, the focus is on common medical issues and rehabilitation concepts that may be useful to primary care physicians in developing a holistic approach to management of spinal cord injury.
The estimated annual incidence of traumatic spinal cord injury in the United States, excluding persons who die before hospital admission, is about 40 cases per million. Between 183,000 and 230,000 persons are living with a spinal cord injury. Although the total number of such patients is not large compared with the number of patients with other neurologic conditions (eg, stroke), the total years of impairment is high because of the young average age at injury (31.2 years). The male-female ratio is slightly higher than 4:1 (1,2).
Motor vehicle crashes account for the largest proportion (36%) of cases of spinal cord injury, followed by acts of violence, falls, and injuries from recreational sports. Since 1973, one notable trend in the etiology of traumatic spinal cord injury has been a proportional decrease in motor vehicle crashes and sports injuries and an increase in acts of violence, primarily gunshot wounds (1,2).
The International Standards for Neurological and Functional Classification of Spinal Cord Injury is the most commonly used classification system among physicians who treat spinal cord injury (3). The neurologic level is defined for each side of the body as the lowest spinal segment with normal motor and sensory function. It is often confused with the skeletal level of injury, which may or may not occur at the same level. A complete injury is defined by the absence of motor or sensory functioning in the lowest sacral segments and is evaluated by testing voluntary anal contraction, perianal sensation, and deep rectal sensation.
Several impairment subscales within this system describe the lowest intact motor level, the lowest intact sensory level, the degree of neurologic impairment below the neurologic level, and the zone of partially preserved innervation. The classification system recognizes five syndromes: central cord, Brown-Séquard, anterior cord, conus medullaris, and cauda equina. The system also includes a disability scale called the Functional Independence Measure.
Medical care in acute spinal cord injury generally involves management of hemodynamic instability, spinal instability, and respiratory insufficiency as well as prevention of a host of potential complications. Care during this phase is best accomplished through concurrent and cooperative management by the trauma team, spine surgeon, and rehabilitation specialist. Because the risk of many complications is reduced and psychological benefits are realized during the inpatient rehabilitation phase, acute medical management should be directed toward expeditious transfer to a regional spinal cord injury rehabilitation unit (4,5).
Hemodynamic instability
Hemodynamic instability often must be treated before definitive spine stabilization can
occur, because spinal cord injury commonly results from multiple trauma. Even in the
absence of bleeding, many patients with acute spinal cord injury are hypotensive owing to
disruption of autonomic cardiac and vasomotor control. In patients with low blood pressure
but no symptoms of hypotension, care should be taken to avoid fluid overload, which can
further compromise respiratory status. Bradycardia or asystole in a small proportion of
patients with cervical spinal cord injury results from unopposed vagal tone. Although
symptoms are usually self-limited, treatment with atropine sulfate may be necessary.
Rarely, if symptomatic bradycardia persists or is refractory to pharmacologic treatment,
cardiac pacing may be indicated.
Spinal instability
Measures to minimize injury to the spinal cord begin with spine immobilization at the
scene of injury when vertebral trauma is suspected. High-dose methylprednisolone sodium
succinate given within hours of the initial trauma in nonpenetrating spinal cord injury is
the current standard of care for preventing secondary injury to the spinal cord (6). When
compromise of the spinal canal is significant, urgent decompression may be accomplished by
traction or may require surgical intervention. Spine stabilization may require internal
fixation. Use of external orthoses, even a halo vest, cannot completely immobilize an
unstable segment.
Timely rehabilitation requires appropriate spine stabilization. Inadequate immobilization of an unstable vertebral segment may lead to progressive neurologic deficit, progressive deformity, malunion, pseudarthrosis, and chronic pain. Prolonged bed rest to allow fusion of an unstable vertebral segment is occasionally necessary when surgical fixation is contraindicated. However, complications such as pneumonia, pressure sores, cardiovascular deconditioning, muscle atrophy, joint contractures, and depression are difficult to avoid.
Respiratory insufficiency
Paralysis of respiratory muscles due to cervical or thoracic level spinal injury may
result in respiratory insufficiency. Onset may be delayed by days owing to eventual
fatigue of partially paralyzed muscles. In some cases, weaning from mechanical ventilation
may be accomplished before transfer to rehabilitation. However, when respiratory muscle
paralysis is significant, gradual weaning through conditioning of partially paralyzed
respiratory muscles may be better pursued concomitantly with other rehabilitation goals.
Excessively rapid weaning inevitably results in recurrent atelectasis, lobar collapse, and
pulmonary infection.
Methods of weaning in this population are controversial because of the lack of controlled clinical trials. However, progressive ventilator-free breathing appears to be more logical from an exercise physiology perspective than the synchronized intermittent minute ventilation or pressure support approach. Both the goals and the approach for weaning must be individualized. Even when use of mechanical ventilation cannot be completely eliminated, periods of ventilator independence or transition to less invasive forms of mechanical ventilation, or both, can reduce the risk of complications and improve quality of life. For selected patients, phrenic nerve pacing may be an option that allows transition to ventilator independence, but it is usually not considered within the first year after spinal cord injury.
Other complications
Secretion management is a crucial part of acute pulmonary care. Retained pulmonary
secretions result from inadequate airflow during coughing and pose a risk of atelectasis
and infection. Copious pulmonary secretions due to reduced sympathetic input and
relatively unopposed vagal tone are common in cervical spinal cord injury. Treatment with
inhaled beta agonists or anticholinergics is often underused and should be considered when
frequent suctioning is required or forced expiratory volume in 1 second improves more than
10% after a test dose. Manually or mechanically assisted cough also should be used more
often to reduce the frequency of deep suctioning, which not only can introduce pathogens
into bronchial structures but also can cause airway irritation, thereby stimulating
further mucus production.
The estimated incidence of deep venous thrombosis in the initial period after spinal cord injury ranges from 47% to 72%, depending on the diagnostic tool used. Data suggest that this complication results from venous stasis as well as hypercoagulability. Autopsy investigations in persons with acute spinal cord injury have shown the death rate due to pulmonary embolism to be as high as 37% (7).
In the absence of contraindications, patients with traumatic spinal cord injury should receive prophylaxis against deep venous thrombosis, including use of compression stockings and sequential compression devices as well as therapy with low-molecular-weight heparin. Inferior vena cava filters may be used when anticoagulation is contraindicated. When a filter has been placed in the absence of contraindications to anticoagulation, heparin administration should be continued to avoid such complications as renal vein thrombosis and postphlebitic syndrome. Diagnosis and treatment of deep venous thrombosis in patients with spinal cord injury are similar to that in able-bodied patients, with one notable exception: The presentation of deep venous thrombosis may be underwhelming in insensate patients; thus, a high index of suspicion is warranted (8).
Pressure sores are common in both acute and chronic spinal cord injury. However, the prevention and management strategies differ at each stage. Before undergoing rehabilitation, patients lack the knowledge and physical resources to prevent skin breakdown. Therefore, the medical team must ensure that pressure is relieved over bony prominences, usually by turning the patient every 2 hours. In addition, the skin must be checked regularly so that pressure over low-grade pressure sores is relieved completely to prevent progression to higher-grade lesions. When pressure sores develop, both local treatment and systemic care are essential. Malnutrition, edema, anemia, and hypoxemia, among other factors, may contribute to delayed wound healing.
The risk of joint contractures increases with prolonged bed rest, especially in the presence of spasticity. Although the best prevention is expeditious transfer to a rehabilitation unit, measures such as appropriate positioning and range of motion during bed rest are important. Splinting has a limited role in prevention of contractures but is generally not effective in regaining lost joint range. Furthermore, splinting requires close monitoring because of the high risk of skin breakdown.
The duration of inpatient rehabilitation varies according to multiple factors, including neurologic level of injury, impairment below the level of injury, comorbidities, the patient's age, premorbid functioning, motivation, social support and, most important, the functional goals for the patient. The rehabilitation specialist individualizes functional goals according to past experience with various degrees of neurologic impairment, the ability to prognosticate neurologic recovery, knowledge of compensatory strategies and adaptive equipment, and the ability to assess complex psychosocial factors.
Table 1 shows the expected functional outcomes at various neurologic levels in patients with motor complete spinal cord injury. In some cases, recovery of a single neurologic level may mean the difference between independence and the need for attendant care. Similarly, psychosocial factors (eg, wheelchair accessibility of the home, a supportive family member) may determine whether the patient can be cared for at home or must be discharged to a long-term care facility. Inpatient treatment is only one step in a structured rehabilitation process that continues for months after hospital discharge. In actuality, rehabilitation is a lifelong process that becomes less structured after about the first year as patients adapt to their impairments, social situations change, aging creates new functional challenges, and new interventions become available (9-11).
| Table 1. Typical functional outcomes for various levels of motor complete spinal cord injury | |||||
| Neurologic level of injury | Respiratory function | Feeding | Dressing | Bowel and bladder function | Mobility |
| C1-C3 | Mechanical ventilation; assisted cough | Dependent | Dependent | Dependent | Power wheelchair with respiratory control; dependent wheelchair transfers |
| C4 | Potential for spontaneous ventilation; assisted cough | Dependent | Dependent | Dependent | Power wheelchair with chin control; dependent transfers |
| C5 | Spontaneous ventilation; assisted cough | Self-feeding with adaptive equipment | Assistance with dressing upper body; dependent with dressing lower body | Dependent | Power wheelchair with hand control; dependent transfers |
| C6 | Spontaneous ventilation | Independent with adaptive equipment | Independent with dressing upper body; assistance with dressing lower body | Potential for independence with orthoses or adaptive equipment | Manual wheelchair for short distances; potential for independent transfers with sliding board |
| C7 | Spontaneous ventilation | Independent with adaptive equipment | Potential for independence with adaptive equipment | Independent | Manual wheelchair; independent transfers |
| C8-T1 | Spontaneous ventilation | Independent | Independent | Independent | Manual wheelchair; independent transfers |
| T2-T10 | Spontaneous ventilation | Independent | Independent | Independent | Manual wheelchair; ambulation with orthoses for exercise |
| T11-L2 | Spontaneous ventilation | Independent | Independent | Independent | Indoor ambulation with orthoses; manual wheelchair for outdoor use |
| L3-S3 | Spontaneous ventilation | Independent | Independent | Independent | Community ambulation with orthoses |
Chronic spinal cord injury affects multiple organ systems to some degree at most levels of neurologic injury. The following discussion highlights issues in management of the most common complications.
Urologic complications
Prevention of common urologic complications requires an understanding of the
pathophysiology of various types of neurogenic bladder that occur in spinal cord injury
and the potential for changes in pathophysiology over time. Regular follow-up evaluation
of both the upper and lower urinary tract is an important part of management.
The goals of bladder management are to maintain continence and prevent complications while minimizing the negative impact on the patient's lifestyle. Upper limb function, sexuality, social support, and personal preference all must be considered. The interactions between medical and nonmedical factors can be complex, and trade-offs are often necessary.
With long-term catheterization, the majority of neurogenic bladders in spinal cord injury eventually become colonized with bacteria. In the absence of symptoms (ie, fever, chills, malaise, autonomic dysreflexia, increased spasticity), a change in urine appearance, foul-smelling urine, catheter clogging, or other evidence of active infection, treatment of bacteriuria is relatively contraindicated because of the potential selection of resistant organisms. However, treatment of infection with stone-forming bacteria in the urine is warranted because of the high incidence of urinary tract calculi that can cause urinary obstruction or bladder irritation, resulting in pain, increased spasticity, or autonomic dysreflexia.
Too often, an indwelling catheter is removed after development of a urinary tract infection in patients with spinal cord injury because the catheter is identified as the cause of the infection. Although a catheter change may be indicated, removal of an indwelling catheter in patients with urethral sphincter dyssynergia and bacteriuria may result in high bladder-filling pressure, ureteral reflux, pyelonephritis, and urosepsis.
Pending the results of urine culture and sensitivity studies, fluoroquinolones are an excellent choice for empirical therapy, because gram-negative organisms are a frequent cause of urinary tract infection. Trimethoprim-sulfamethoxazole (Bactrim, Septra, Sulfatrim) is a reasonable alternative in outpatients, because Pseudomonas species, although frequent colonizers, are rarely pathogenic in chronic spinal cord injury (12).
Gastrointestinal complications
Gastrointestinal problems account for a significant proportion of patient-care hours in
spinal cord injury. The importance of proper neurogenic bowel management cannot be
overemphasized because of its effect on patients' interpersonal, sexual, and vocational
roles. The goals of bowel management are to maintain fecal continence and regulate
defecation while minimizing the negative impact on the patient's lifestyle.
Because the lower motoneuron cell bodies of the bowel reside in the nerve plexuses of the bowel wall, reflex activity and local control usually remain intact. Peristalsis continues, but transit time increases. In suprasacral injuries, anal sphincter tone is typically adequate to maintain continence, provided that defecation can be triggered through reflex mechanisms on a regular basis. These mechanisms include use of digital stimulation of the anal sphincter, suppositories that cause local irritation of the rectal mucosa, and oral agents, including foods, that may stimulate defecation. Regulation of stool consistency is crucial and is best achieved through diet, but the aid of oral stool softeners, bulk laxatives, or both, may be needed.
Autonomic dysreflexia
This potentially life-threatening condition occurs in some patients with spinal cord
injury above the midthoracic level. Unopposed sympathetic outflow results from nociceptive
input below the injury level. Hypertension with systolic blood pressures up to 250 mm Hg,
if left untreated, may result in intracranial hemorrhage. Other signs and symptoms include
tachycardia or reflex bradycardia, headache, facial flushing, urticaria, and diaphoresis.
The mainstay of treatment is identification and removal of the nociceptive input. Because the majority of patients are insensate, this process consists of systematically ruling out the most common causes. If the source cannot be identified and removed immediately, temporizing measures to prevent intracranial bleeding must be undertaken. Elevation of the head and treatment with a short-acting antihypertensive agent, such as nitroglycerin ointment (Nitro-Bid, Nitrol), are reasonable choices. Longer-acting antihypertensive agents often lead to hypotension once the source of nociception is removed. The bladder, bowel, and skin are the most common sources. Therefore, the first steps in diagnosis are to rule out bladder distention, urinary tract infection, fecal impaction, and skin lesions.
When the source is identified but cannot be removed immediately, treatment with local anesthesia, oral analgesics, or longer-acting antihypertensive medications may be indicated. Long-term prophylactic treatment is occasionally required for patients who are prone to frequent recurrent episodes (13).
Spasticity
Spasticity is common in spinal cord injury and has both potentially beneficial and
detrimental effects. In some patients, spasticity may help to maintain muscle bulk, reduce
the risk of deep venous thrombosis, or aid in transfers or ambulation. In others, it may
(1) interfere with sitting, transfers, or ambulation, (2) predispose to joint contractures
or skin breakdown, or (3) cause pain. Therefore, management must be individualized on the
basis of function. Unlike in stroke, baclofen (Lioresal) is a first-line pharmacologic
intervention in spinal cord injury. Treatment of spasticity is an important component of
managing many upper motoneuron disorders.
Skin breakdown
Skin care is an essential component of health maintenance in patients with spinal cord
injury. The consequences of inadequate attention to this aspect of care can be
devastating. Too often, recurrent skin breakdown results in limb amputation,
osteomyelitis, sepsis, or death. As with many health maintenance issues, an ounce of
prevention is worth a pound of cure. The key to prevention is patient education.
The integumentary system is at risk in patients with spinal cord injury because of lack of protective sensation, decreased soft-tissue bulk, histologic skin changes, impaired mobility, and sometimes urinary or fecal incontinence. High-risk areas are bony prominences that receive high pressure during sitting or recumbency, such as the sacrum, ischial tuberosities, greater trochanters, heels, and occiput. The mainstay of prophylaxis is regular weight shifting, which may be performed by the patient with or without assistive equipment or by an attendant. In addition, the skin must be checked daily to allow early identification and treatment of pressure sores and to avoid progression to overt ulceration.
Understanding of the physiologically complex mechanisms underlying spinal cord injury and neurologic recovery has advanced tremendously over the past decade. The neurologic deficits clearly result not only from mechanical injury and disruption of blood flow to the spinal cord but also from secondary biochemical injury that is partially preventable with appropriate treatment (6). In addition to the injury to the spinal cord, root level injury is commonly seen near the level of skeletal injury.
Documented mechanisms of neurologic recovery include remyelinization of neurapraxic neurons, axonal regeneration, and peripheral sprouting of intact neurons (14). Many other mechanisms have been postulated, including cortical reorganization and ephaptic nerve transmission. Clinical data suggest that although the predominance of motor and sensory recovery takes place within the first several months after injury, gains can continue for more than 2 years in some cases. Even patients with a complete spinal cord injury may regain one or two spinal levels (4).
Research in spinal cord injury has reached an exciting stage in its evolution. Many areas of research involving central nervous system (CNS) pathology have shown that the CNS is more capable of neural plasticity than previously thought. Animal studies involving spinal cord grafting have shown that partial motor recovery after spinal cord transection is possible. Functional electrical stimulation (FES) involves electrical stimulation of paralyzed muscles, causing them to contract in a coordinated manner to produce functional movement. The Food and Drug Administration recently approved a surgically implanted, FES-based neuroprosthesis that restores hand grasp to patients with C5 and C6 tetraplegia. Ongoing FES research includes applications for improving upper and lower limb function, bladder control, spasticity, fertility, and skin care. Furthermore, such research suggests that peripheral mechanisms can be employed to influence CNS plasticity in upper motoneuron paralysis.
Management of spinal cord injury is a continuum from initial emergency medical treatment to lifelong medical and rehabilitative care. Because improved care for patients with spinal cord injury has resulted in increased longevity, cardiovascular and musculoskeletal issues have become increasingly prevalent in this aging population. Given the limited number of medical centers with special expertise in spinal cord injury, interested primary care physicians potentially have a valuable role in providing care for this population. The wide range of medical issues and the functional goals of patients with spinal cord injury are interdependent. Thus, optimal care requires a coordinated effort between primary care physicians and rehabilitation specialists.
The appendix is a narrow, worm-shaped pouch of tissue, usually less than 1 in. wide and 3 to 4 in. long, that extends from the large intestine. Its purpose is not known, but some researchers think it may help trigger the immune system to fight disease. Getting along without one is no problem--which is a good thing, since the most common reason for abdominal surgery is removal of an inflamed appendix.
What causes appendicitis?
The appendix is a "dead end." When something (such as a hard pellet of stool, a
foreign body, a seed or other piece of undigested food) blocks the only opening from the
colon, inflammation soon sets in. In other instances, infection from bacteria in the
intestinal tract causes swelling that gradually closes the opening. There is no way to
predict or prevent appendicitis. It can occur in anyone. However, it is most common in
adolescents and young adults and rare in children under age 2.
Is appendicitis serious?
When it's not treated, it can be very serious. A completely obstructed appendix becomes
more and more inflamed and eventually bursts, casting out its infected contents. The
entire lining of the abdominal cavity may become contaminated, causing a life-threatening
condition called peritonitis.
What are the symptoms of appendicitis?
Symptoms are usually vague at first, but an early one may be dull, continuous pain in the
area around the navel. As inflammation worsens, several of the following symptoms may be
noted:
How is it diagnosed?
Appendicitis can be hard to diagnose because symptoms often resemble those of bladder
infection, kidney stones, inflammation of the colon or stomach, or ovarian cysts. Physical
examination is usually the first step. By pressing on the right lower part of the abdomen,
the doctor may be able to feel a hardened appendix. By letting up quickly, he or she may
notice rebound tenderness (pain that is worse after pressure is relieved), which is a sign
of appendicitis. Rectal and pelvic examinations often are done to rule out other problems.
Usually, a blood sample is tested for a high white blood cell count indicating infection, and a urine sample is tested for urinary tract infection, which can be confused with appendicitis. An x-ray film may be taken to look for an obstruction, or a thin, lighted telescope called a laparoscope may be inserted into the abdomen to examine the area.
What should I do if I suspect appendicitis?
If you think a youngster may have appendicitis, keep a record of his or her temperature,
taken every 2 hours. If there is a sudden rise, pain gets worse, or the child starts
vomiting or has blood in the stool, call your doctor right away. Whether you suspect the
problem in yourself or someone else, remember the following tips to avoid making the
situation worse:
How is appendicitis treated?
An inflamed appendix has to be removed, but the operation is simple and usually free of
complications. The patient is given a general anesthetic and sometimes an antibiotic to
fight infection. A 3- or 4-in. incision is made in the lower abdomen, or if a laparoscope
is used for the operation, the incision is even smaller. The appendix is cut away, and the
wound in the intestine is closed and cauterized. If there is any fluid or pus present, it
is suctioned away, and the abdominal incision is closed with sutures. Patients usually
stay in the hospital 3 or 4 days.
After returning home, patients may want to take an over-the-counter pain reliever for a few days. If pain gets worse or if swelling, redness, drainage, or bleeding in the surgical area begins, the doctor should be consulted. Additional symptoms that should be reported are nausea, vomiting, abdominal swelling, and signs of infection (such as headache, muscle aches, dizziness, fever). Most people are back to their normal activities in about 3 weeks.
Still no clear evidence to link specific beverages to specific cancers
Throughout the world the connection between alcohol and mortality in particular deaths from coronary diseases is described as a U shaped curve, with the highest mortalities among non-drinkers and heavy drinkers and the lowest among moderate drinkers. Recently the discussion has been extended to consider whether there might be a difference in mortality from coronary disease depending on the type of alcohol consumed. Some have suggested that wine in moderate quantities reduces mortality, but in 1997 White concluded, "No real evidence exists that wine is better than beer and spirits."1 Similar questions have also been raised about the possible carcinogenic effects of alcohol and whether these too are related to different types of alcohol.
The most well researched associations between alcohol and cancer are in breast cancer in women, and in cancer of the mouth, oesophagus, pharynx, larynx, and liver,2 but other types of cancer have also been discussed: colorectal, bladder, and pancreatic cancer. Recent studies have investigated whether the incidence of and mortality from specific cancers could depend on alcohol type; as with coronary disease it is difficult to point to certain evidence that wine is less dangerous than beer or spirits or vice versa. In a study of pancreatic cancer Partanen et al found that consumption of distilled drinks did not increase the risk whereas heavy drinking of wine and beer did.3 In contrast, Launoy et al showed that the risk of oesophageal cancer was due to variations in local drinking behaviour, with hot spirits and beer carrying the highest risk. Even after adjustment for consumption of all other alcoholic drinks, consumption of hot calvados explained almost half of the variation in the study.4 The study by Grønbæk et al in this issue further explores the relation between upper digestive tract cancers and alcohol.5 They found that a moderate intake of wine did not increase the risk of these cancers, whereas an equal intake of beer or spirits increased the risk considerably. In 1997 Viel et al showed that the incidence of breast cancer increased with increased alcohol consumption, though this increased risk seemed to be restricted to red wine consumption.6
Some studies suggest that alcohol may cause other types of cancer. For example, in a case-control study of the rare cancer of major salivary glands Horn-Ross et al estimated an odds ratio of 2.5 among male heavy alcohol consumers but not among women.7 However, Muscat and Wynder did not find that alcohol increased the risk of salivary gland cancer.8 A Polish study suggested that strong spirits were a risk factor for lung cancer among women: they found a higher incidence of lung cancer among female vodka drinkers.9
There may be many explanations for these results. Some may be due to coincidence, since only a few studies had an a priori hypothesis. Other findings, such as the association between vodka and lung cancer, might be caused by residual confounding even after adjustment in the multivariate models.
Several hypotheses exist about the mechanism whereby alcohol consumption increases the risk of cancer. Some are based on studies of biomarkers. Strom et al found that alcohol induced sister chromatid exchange and a significantly higher frequency of spontaneous breaks of the chromosomes.10 More recently, however, Spitz et al concluded that mutagen sensitivity is an independent marker of cancer risk not affected by other known risk factors such as smoking or alcohol consumption.11
The conclusion so far must be that an association between alcohol and cancer certainly exists for most of the above mentioned cancers. It is still not possible, however, to distinguish between the risk factors associated with different types of beverage, and if a difference should emerge we do not yet understand the underlying mechanism.
Svend Sabroe
Department of Epidemiology and Social Medicine. University of Aarhus, 8000 Århus C, Denmark