This study helps shed more light on LBP that motivates patients to seek medical care. More than 40% of LBP patient visits were provided by family (general) medicine physicians, comprising one of every 17 patient visits for this specialty. However, orthopedic surgeons provided a slightly greater percentage of LBP visits as part of their specialty case mix (1 of every 16 patient visits). Further, a substantial percentage of LBP patient visits (45%) were provided by non-primary care physicians. During 2003–2004 there were more than twice as many patient visits annually for LBP than reported in the 1990 NAMCS; however, the percentage of patient visits attributed to LBP (3%) and the percentage of LBP patient visits provided by primary care physicians (55%) remained remarkably similar to what was reported in the 1980s and early 1990s . Non-primary care services generally are considered inappropriate for patients with non-specific LBP . Thus, this study suggests that the recommended shift to primary care physicians for medical management of non-specific LBP has not occurred over the past two decades.
Injuries were the strongest risk factor associated with LBP patient visits (OR, 3.38; 95%, 2.75–4.14). There were also characteristic patterns of LBP patient visits according to age and geographic region. Patients other than those 25–44 years of age were less likely to seek medical care for LBP. Unavailable, and therefore uncontrolled, variables that could potentially explain the observed age distribution of LBP patient visits include occupational risk factors such as manual handling of materials, bending and twisting, whole-body vibration, and lifting for more than three-fourths of the work day . Patient visits for LBP were less likely to occur in the Midwest and South than in the Northeast (OR, 0.56; 95% CI, 0.40–0.77 for each contrast). It is unclear if these geographic findings reflect the epidemiology of LBP in the United States or if they are confounded by other uncontrolled variables. While relatively little is known about risk factors in the transition from acute to chronic LBP, this study suggests that injuries are not associated with progression of LBP.
Osteopathic physicians were more likely than allopathic physicians to provide medical care during LBP patient visits (OR, 2.61; 95% CI, 1.75–3.92). The physician specialties most likely to provide LBP patient visits were family (general) medicine, internal medicine, neurology, and orthopedic surgery. These findings are consistent with previous studies [10, 22]. There was an even stronger association between osteopathic physicians and chronic LBP patient visits (OR, 4.39; 95% CI, 2.47–7.80). However, physician specialists in family (general) medicine, internal medicine, and neurology were not more likely than other physician specialists to provide chronic LBP patient visits. These findings, coupled with the greater use of shared physician care in chronic LBP (OR, 2.11; 95% CI, 1.30–3.44), suggest that osteopathic physicians are often used to complement the conventional medical management of chronic LBP by providing spinal manipulation.
At least eleven national clinical guidelines for LBP medical management in the primary care setting were published between 1994 and 2000 . An updated review of national clinical guidelines summarized recommendations according to LBP chronicity . For acute LBP, radiographs were not considered useful for diagnosis of non-specific LBP. Recommended treatments included advising patients to remain active (although back-specific exercises were not considered effective), and ordering paracetamol or NSAIDs (muscle relaxants or narcotic analgesics may be considered as well). In contrast to acute LBP, few guidelines existed for the medical management of chronic LBP. Recently, however, European guidelines have been established for the management of chronic non-specific LBP . These guidelines do not recommend radiographs or other diagnostic imaging tests unless a specific cause is strongly suspected. They recommend brief educational interventions (specifically including supervised exercise therapy), cognitive behavioral therapy, and short-term use of NSAIDs or weak narcotic analgesics for pain relief. They generally do not recommend physical therapies (although spinal manipulation may be considered) or surgery (unless all other recommended conservative treatments have been tried and failed over a period of at least two years).
This study suggests that cognitive behavioral therapy (as proxied by mental health counseling) may be under-utilized in the medical management of chronic LBP (6% of patient visits), which is often characterized by depression and somatization . Nonsteroidal anti-inflammatory drugs were the most commonly used drugs for acute LBP (49% of patient visits); however, they were less likely to be used for chronic LBP (28% of patient visits). Non-narcotic analgesics were infrequently used for either acute or chronic LBP (6% of patient visits overall). The reported percentage of chronic back patients prescribed narcotic analgesics varies widely, from 3% to 66%, based on the treatment setting . The present study found the relevant percentage to be 28%. Almost two million surgical procedures (about one million annually) were ordered, scheduled, or performed during primary LBP patient visits. Not surprisingly, surgical procedures were more frequently associated with chronic LBP patient visits compared with acute LBP patient visits (OR, 12.62; 95% CI, 3.18–50.07). Together, the findings of this study reinforce the caricature of LBP medical care in the United States as being overspecialized, overinvasive, and overexpensive .
The medical management of LBP varies substantially between practitioners and countries [24, 28]. Differences were observed in this study with regard to type of physician provider, physician specialty, and geographic region. Osteopathic physicians were less likely than allopathic physicians to order NSAIDs for LBP (OR, 0.43; 95% CI, 0.24–0.76). This is consistent with the theory that osteopathic physicians are less likely to prescribe drugs for LBP because they may use spinal manipulation as an alternative to drugs . Previous studies including an analysis of older NAMCS data  and a randomized controlled trial  have provided evidence to help support this theory. Family (general) medicine physicians were less likely to provide exercise counseling, but were more likely to order drugs for LBP. Time constraints during patient visits, particularly in a managed care environment, represent a possible explanation for the latter findings . Patient counseling was less often provided outside the Northeast in this study. A strong predictor of patient counseling, including exercise counseling, was having had an injury as the reason for seeking medical care for LBP (OR, 2.38; 95% CI, 1.50–3.77).
Although this study involved a large, nationally representative sample of patient visits for ambulatory medical care in the United States, there are several limitations of this study that should be noted. The study involved the epidemiology and medical management of LBP that was of a magnitude sufficient to prompt patients to visit physician offices for ambulatory medical care. Further, the study was limited by the NAMCS patient record form to patient visits in which LBP was among the three most important reasons for seeking medical care. Thus, this may more properly be considered a study of the epidemiology and medical management of clinically significant LBP in the ambulatory medical care environment. Although the measurement of incidence or prevalence rates was not an objective of the study, all patient visits in which LBP was recorded as a reason for seeking medical care were included in the epidemiological analyses to capture the maximal number of incident or prevalent LBP cases and thereby to provide more precise statistical estimates. Nevertheless, similar results were observed in the epidemiological analyses when only primary LBP patient visits were included (Table 1). The medical management analyses, however, were limited to only primary LBP patient visits to avoid potential confounding by other more important reasons for seeking medical care.
Simplifying assumptions were made in certain analyses because of limitations inherent in the NAMCS patient record form. Patient visits attributed to back (RFV code 1905) and low back (RFV code 1910) symptoms were combined because there were no substantive differences in the characteristics associated with these reported reasons for seeking medical care (Table 1). All patient visits attributed to back symptoms were assumed to involve back pain, although 4% of such visits involved unspecified back symptoms and another 2% involved such other back symptoms as cramps, contractures, spasms, limitation of movement or stiffness, or weakness. Similarly, with regard to etiology, all of the 19.7 million patient visits in which an IPA was reported (using a dichotomous patient record form item) were assumed to involve an injury, although in the subset of 8.4 million patient visits in which the specific IPA was described, up to 11% may have involved iatrogenic, environmental, or other etiologies. The elements of LBP medical management were assessed with survey items that asked whether the relevant element was "ordered, scheduled, or performed." However, it was impossible to confirm whether the reported elements actually occurred within the relevant patient visit or were eventually performed by the reporting physician.
Several analyses yielded imprecise results because they were based on less than 30 NAMCS patient visits or because the SE was greater than 30% of the NPE. Most often this occurred with less common characteristics (internal medicine physicians) or elements of LBP medical management (weight reduction and mental health counseling, physiotherapy, and surgical procedures), or in stratified (subgroup) analyses. Thus, racial minority groups were combined in a "non-White" group to partially overcome this limitation. Hospital admission could not be studied as an element of LBP medical management because of the limited number of observations.
In conclusion, this study found that the percentage of LBP visits provided by primary care physicians in the United States remains suboptimal. Medical management of LBP, particularly chronic LBP, appears to over-utilize surgery relative to more conservative measures such as patient counseling, non-narcotic analgesics, and other drug therapies. Osteopathic physicians are more likely to provide LBP care, and less likely to use NSAIDs during such visits, than their allopathic counterparts. In general, LBP medical management does not appear to be in accord with evidence-based guidelines.