Topic: Sciatica and Low Back Pain: Does Physical Therapy (Read 951 times)
member is offline
Sciatica and Low Back Pain: Does Physical Therapy
« Thread started on: Jan 22nd, 2008, 7:10pm »
Sciatica and Low Back Pain: Does Physical Therapy Provide Long-Term Benefits? A Best Evidence Review CME/CE
Jacqueline A. Hart, MD
Best Evidence Reference
Cost-Effectiveness of Physical Therapy and General Practitioner Care for Sciatica
Luijsterburg PA, Lamers LM, Verhagen AP, et al.
This study was selected from Medscape Best Evidence, which uses the McMaster Online Rating of Evidence System. Of a possible top score of 7, this study was ranked as 6 for newsworthiness and 5 for relevance by clinicians who used this system.
The incidence of lumbosacral radicular syndrome (LRS) or sciatica in the United States has been estimated at 1% to 10%. The authors of the current study estimate that the incidence of LRS is 0.5% in The Netherlands. The cost of caring for those with sciatica, therefore, is astronomical -- an estimated 1.18 billion euros in direct and indirect costs each year in The Netherlands and more than $50 billion annually in the United States. Luijsterburg and colleagues at the Erasmus Medical Center in Rotterdam, The Netherlands, sought to determine whether physical therapy (PT) in addition to conservative treatment consisting of care by a general practitioner (GP) is cost-effective as measured by direct and indirect costs.
Care by the GP followed specific clinical guidelines, which included information and advice about LRS as well as prescription pain medication if needed. Those randomized to PT, using a computerized randomization process, received additional information and advice about LRS as well as PT. The physical therapists essentially coached and guided these patients back to activities as usual. Massage, manipulation techniques, ultrasound therapy, and electrotherapy were specifically not involved as part of the PT offered to subjects in the study.
Inclusion criteria for subjects, who were recruited between May 2003 and November 2004, were as follows:
Radiating pain to the leg below the knee;
Severity of symptoms above 3 on a scale from 0 (no pain) to 10;
Duration of symptoms less than 6 weeks;
Age between 18 and 65 years; and
Pain with coughing, sneezing, or straining; decreased muscle strength in the affected leg; sensory deficits in that leg; decreased reflex activity in that leg; or positive straight leg raising test.
Subjects were excluded for the following reasons:
Radiating symptoms had also occurred at another time in the preceding 6 months;
Back surgery had been performed in the previous 3 years;
Epidural injection treatments had taken place in the past;
Patient was pregnant;
Comorbidity limited overall well-being;
Indication for surgery was present at baseline (eg, progression of paresis or presence of cauda equina syndrome); and
Patient was expected to be lost to follow-up (eg, moving from Rotterdam).
A total of 135 subjects were randomized to receive PT along with GP care (n = 67) vs GP care alone (n = 68). Although the subject list was developed by an independent person running the computerized program using a concealed process, the allocated treatment could not be blinded in terms of the recipients of care or the clinicians (GPs or physical therapists) treating them. Baseline characteristics and drop-out rates were similar between groups, with 87% of participants (117 subjects in total) completing the study at 1-year follow-up.
The primary outcome measure was global perceived effect (GPE), assessed on a 7-point scale from 1 (completely recovered) to 7 (vastly worsened) and reported as percentage improved (defined as completely recovered and much improved) vs percentage not improved (defined as slightly improved, not changed, slightly worsened, much worsened, and worse than ever). Secondary outcome measure was EuroQol, a standardized tool developed in Rotterdam, the results of which can be converted to calculate quality-adjusted life years (QALYs).
Outcome measures and costs were assessed at baseline and at 3, 6, 12, and 52 weeks after randomization. Costs included the following:
Manual therapy visits;
Consultation with specialists;
Help with housekeeping (direct nonhealthcare costs); and
Absence from paid work (indirect costs).
The incremental cost-effectiveness ratio (ICER) was evaluated for both groups of subjects using total costs and direct costs only. Confidence intervals (CI) for the ICER were calculated, and cost-effectiveness acceptability curves were constructed. In terms of the primary outcome, 79% of subjects receiving PT vs 56% of the GP-only group reported improvement by GPE; this was significantly different (relative risk: 1.4; 95% CI: 1.1;1.8). At 1-year follow up, there was no statistical difference between groups (PT vs GP-only) in percentage of visits to neurosurgeon or orthopedic surgeon (8% vs 10%) or percentage who underwent surgery for LRS (6% vs 4%). There was also no statistical difference in QALYs between the 2 groups at the end of the year; in fact, the only time point that showed a statistical difference in QALYs was at 6 weeks, which favored the GP-only group.
In terms of cost analysis, the ICER was calculated based on GPE score for direct costs only and for total costs, which were 837 and 6224 euros, respectively. This means that for every patient in the PT group who reported improvement by GPE, extra direct costs were 837 euros and extra total costs (ie, including both direct and indirect costs) were 6224 euros, the bulk of which was from loss in productivity.
Put simply, the 6224 euros that PT added to the cost for improvement in GPE, with no difference in QALYs, hardly seems worth it. As the researchers point out, this study even raises the possibility that PT contributes to lost productivity because attending PT appointments could lead to absence from work and contribute to production losses. Should the very common clinical practice of PT referral, therefore, be continued? The researchers conclude that it is not cost-effective to continue this practice.
Of concern, however, is that most healing modalities used by physical therapists were specifically avoided in the study design. The physical therapists were not allowed to use massage therapy, manual manipulation, electrotherapy, or ultrasound -- one or more of these components is typically a very important part of a physical therapist's treatment repertoire. Providing only additional education and some exercise therapy then seems extremely limited and not differentiated enough from a visit to a GP to demonstrate substantial cost benefit for PT. Eliminating all PT components other than education and exercise could easily explain the extreme imbalance seen in this study, in which costs were high but benefits were limited. Quite literally, the physical therapists' hands were tied.
The controversy regarding the value of PT for sciatica and other causes of acute low back pain (LBP) has been debated for a long time. Even guidelines for how to treat acute LBP vary from country to country. In an excellent head-to-head comparison of assess/advise/treat vs assess/advise/wait, Wand and colleagues randomized 102 eligible subjects to compare these 2 models and determined that at 6 weeks there were significant differences in favor of the assess/advise/treat group over the assess/advise/wait group on a host of primary and secondary measurements of:
Quality of life;
Social functioning; and
Neither pain nor disability was significantly different between the groups at long-term follow-up but certain important differences persisted at 3 and at 6 months in favor of the assess/advise/treat group over the assess/advise/wait group. Measurements favoring assess/advise/treat were:
Quality of life;
Feelings of anxiety and depression; and
Less interference of emotional difficulties in daily activities.
In spite of these findings, Dutch and Australian guidelines recommend waiting and not prescribing PT, whereas American and British guidelines advise immediate PT. Of interest, the study under review, conducted in the Netherlands, supports the Dutch guidelines, and the study by Wand and colleagues, which was conducted in Australia, supports the British guidelines of advising active PT for treatment of acute LBP, including sciatica, early on.
In addition, 2 red flags in the current Dutch trial might make one hesitant to give up on early PT for LRS and other causes of LBP. As discussed above, the design of the current study did not allow for PT modalities other than exercise. Although similarly, Wand and associates did not allow electrotherapy or traction but they did allow manual therapy, including low-velocity mobilization techniques and high-velocity manipulation at the therapist's discretion. The study by Wand and coworkers more accurately reflected PT as it is actually practiced. In addition, the Luijsterburg trial did not consider measurements of anxiety, depression, and other emotional difficulties secondary to the pain and disability that may be interfering with activities of daily living. What are the expenses associated with those common consequences of acute turned chronic LBP and sciatica? Are these costs accounted for in the Dutch cost-effectiveness trial?
With that said, the study by Luijsterburg and colleagues suggests that as primary care clinicians, we should think about broader implications when choosing whether PT is right for an individual patient with LRS or other forms of acute LBP and not automatically prescribe it as part of the treatment plan.