OUP user menu

Controlled Zygapophysial Joint Blocks: The Travesty of Cost-Effectiveness

Nikolai Bogduk MD, PhD, DSc, Scott Holmes PhD
DOI: http://dx.doi.org/10.1046/j.1526-4637.2000.99104.x 24-34 First published online: 1 March 2000

Abstract

Objective. The aim of this study was to develop equations by which the costs could be compared of various models of performing diagnostic blocks for spinal pain.

Design. Algorithms were elaborated describing different strategies for the diagnosis of cervical or lumbar zygapophysial joint pain using placebo-controlled diagnostic blocks, comparative local anaesthetic blocks, or no control blocks, and its treatment with radiofrequency neurotomy. For each step in each algorithm cost functions were applied. Summary equations were derived that allowed the cost of the algorithms to be compared algebraically. A selection of costs were substituted for the unknown variables in the equations in order to illustrate the cost-effectiveness of different algorithms under Australian and US conditions.

Results. The equations indicated that cost-effectiveness was critically dependent on the ratio between the cost of treatment and the cost of a diagnostic block. For cervical zygapophysial joint pain, reimbursements discourage best practice, both in Australia and in the United States, by rendering the use of controlled blocks more expensive than no controls. For lumbar zygapophysial joint pain, controlled blocks are cost-effective under Australian fee schedules, and under some but not all American schedules. In the name of cost-effectiveness, the US fee structure encourages presumptive therapy without regard to diagnosis, but ignores the ethical and logistic consequences of inordinately high failure rates of therapy when a diagnosis is not established using controlled blocks.

Conclusions. Best practice, using placebo-controlled diagnostic blocks before neurosurgical therapy of zygapophysial joint pain, is not encouraged and rewarded in the United States. In Australia it is compensated only in the context of lumbar zygapophysial joint pain. In the interests of short-term financial savings, the US fee structure sacrifices the majority of patients to failed treatment because of lack of proper diagnosis. Clinical absurdity, rather than evidence-based, best practice is encouraged.

  • Back pain
  • Neck pain
  • Diagnosis
  • Treatment
  • Cost-effectiveness
  • Reimbursement
  • Zygapophysial joint

Systematic studies by investigators in Australia and the United States have demonstrated the validity and utility of diagnostic blocks of the zygapophysial joints in the management of spinal pain. Controlled studies have shown that among patients with chronic back pain the prevalence of zygapophysial joint pain is of the order of 15%[1] and increases to 40% in some populations[2]. Double-blind, controlled studies have shown that cervical zygapophysial joint pain accounts for 50% of patients with chronic neck pain after whiplash[3,4]. Moreover, open[5,6] and controlled trials[7,8] have shown that both lumbar and cervical zygapophysial joint pain can be successfully treated by percutaneous radiofrequency neurotomy.

However, zygapophysial joint pain cannot be diagnosed on the basis of history or clinical examination[1,2,9] or medical imaging[10]. Diagnostic blocks constitute the only means for diagnosis, yet they are subject to false-positive responses. Patients may respond positively to a single diagnostic block for reasons other than the action of the local anaesthetic administered. Formal studies have shown that, in the case of lumbar zygapophysial joint blocks, of every 100 patients who report a positive response to a first block, only 31 prove to be true-positive when controlled blocks are used[11]. In the case of single, cervical zygapophysial joint blocks, 78% of patients prove true-positive but 27% prove false-positive[12]. The difference in false-positive rates reflects the difference in prevalence of lumbar and cervical zygapophysial joint pain; lumbar zygapophysial joint pain is far less common than cervical zygapophysial joint pain and, therefore, associated with a far greater false-positive rate.

Therefore, single, diagnostic blocks lead to incorrect diagnoses in a substantial proportion of cases. For this reason, the research literature has advocated the use of controlled blocks in each and every case to confirm the response and to prevent false diagnoses from being made. This call, however, has been met with less than full compliance.

Although not published in the formal literature, several reasons for not complying with controlled blocks have frequently been raised in discussions at conferences at which controlled blocks have been advocated. Paramount among these reasons have been claims that controlled blocks are not cost-effective. Some, if not many, practitioners of diagnostic blocks claim that either it is not efficient to perform controlled blocks or that insurers are not willing to pay for what they perceive to be additional but superfluous procedures. They consider it more expedient and cost-effective to offer patients treatment based on the results of single diagnostic blocks, or even no blocks at all in some instances.

These arguments conflict with what research defines as best practice. In principle, one would expect that a diagnosis should be made as accurately as possible (within reason). Moreover, invasive and destructive treatment should be applied only if the diagnosis is reasonably certain and not arbitrarily in the hope that it might work. In the context of zygapophysial joint pain, the research literature warns that single blocks are not reliable. For an accurate diagnosis to be made, controlled blocks are mandatory. If an accurate diagnosis is not made, then treatment is destined to fail in those patients with an incorrect diagnosis. Although limiting investigations to single, diagnostic blocks may seem expedient, the costs of failed treatment may outweigh any apparent savings achieved by curtailing costs of investigation. This has significant financial implications for those who underwrite treatment costs.

This study considers the cost-effectiveness of diagnostic blocks and provides data upon which the claims of advocates of single blocks could be objectively assessed. The study models the cost implications of diagnostic decisions based on diagnostic blocks, comparing the effects of single diagnostic blocks with those of controlled diagnostic blocks.

Methods

Models were constructed by listing the steps through which a patient might pass in the investigation of their spinal pain using zygapophysial joint blocks. For each step, the likelihood of possible results was included using available epidemiologic data. For positive responses to diagnostic blocks, possible treatment options were added together with the likelihood of favorable responses and failures. The treatment for a positive response to blocks was taken to be radiofrequency neurotomy because this is the only validated treatment for zygapophysial joint pain[7,8]. Other treatment options, such as intraarticular steroids, were not considered because these have been refuted by controlled trials[13,14]. At this stage, no other treatment has been tested for zygapophysial joint pain.

For the purpose of constructing the models presented, it was assumed that patients whose diagnosis was not confirmed by placebo-controlled diagnostic blocks, but who nonetheless underwent treatment, would ultimately fail to obtain relief from that treatment. Placebo responses to treatment were not factored into the cost analysis because it has been shown that radiofrequency neurotomy carries a minor placebo effect that is not long lasting and is tantamount to failure[7,8].

Separate models were constructed for cervical pain and for lumbar pain because the biological data for these 2 index conditions are different. Separate models were constructed for the single block process and the controlled block process using either comparative local anaesthetic blocks[15] or placebo-controlled blocks[4,7,16]. In all models the cost analysis was applied to all steps after an initial diagnostic block on the grounds that an initial diagnostic block is required for all models to identify putatively symptomatic patients. Thus, the analysis undertaken is independent of the prevalence of zygapophysial joint pain; it starts with patients believed to have zygapophysial joint pain on the basis of an initial diagnostic block. Whether patients should undergo diagnostic blocks in the first instance and whether these blocks are cost-effective is another and separate issue. This study focused only on the comparative costs that follow once an initial block proves positive.

For each step in each model, available cost data were applied. In the first instance, these costs were treated as an unknown to generate a set of equations for calculating total costs. Subsequently, representative costs for each step were introduced to solve the equations. The first solutions focused purely on medical costs: the immediate costs of performing the diagnostic and therapeutic procedures. Other costs, such as the costs of failed treatment and opportunity costs, were not included, but the equations were developed in such a form that these additional costs could subsequently be factored into the solutions, if required.

To provide concrete examples, the equations were solved by substituting the costs of diagnostic blocks and radiofrequency neurotomy as they apply in Australia and in the United States. The Australian figures were drawn from 2 schedules that operate in the country: (1) Medicare fees, which are fees that the government recognizes for patients treated under the public health system; and (2) Australian Medical Association (AMA) fees, which are fees recommended by the AMA, and which typically apply to private patients and patients whose care is paid for by worker's compensation and third-party insurers.

Difficulties arose in obtaining representative fees for the United States. Different fees apply in different states and under different systems. There is no universal schedule of fees. Moreover, when asked to provide an indication of their fees, American colleagues were reluctant to divulge them; some claimed that to do so would breach antitrust laws. Consequently, the US fees used were those provided by 2 colleagues in 2 different states, who were prepared to reveal their fees but who wished their names and addresses to remain anonymous.

Results

The most elaborate model for diagnosing and treating zygapophysial joint pain involves double blocks with different local anaesthetic agents and a placebo block with normal saline. This has been described and applied for the investigation and treatment of cervical zygapophysial joint pain[7,14,16]. In this study, we refer to this elaborate model as model A, which is shown inFigure 1. It involves determining, in the first instance, whether a joint is putatively symptomatic. Patients with a positive response to a first block undergo a control block and a placebo block. In actual practice the control block and the placebo block are randomized, but for our purposes they are shown in sequence inFigure 1 because the nature of the information drawn from each block is different.

Figure 1

Model A. Flow chart of the process and outcomes of placebo-controlled, local anaesthetic diagnostic blocks for cervical zygapophysial joint pain. *Patients negative to comparative blocks are by definition negative, regardless of their response to placebo.

Comparative local anaesthetic blocks identify 3 types of responses[15]: (1) negative, in which patients fail to obtain relief when a second local anaesthetic is used; (2) discordant, in which patients obtain short-lasting relief when a long-lasting local anaesthetic is used, and long-lasting relief when a short-acting agent is used; and (3) concordant, in which patients respond physiologically in accordance with the expected duration of action of the agents used. Two studies[12,15] have shown that, in the investigation of cervical zygapophysial joint pain, the proportions of patients exhibiting these responses after an initial positive response to blocks are approximately 0.02, 0.22, and 0.76, respectively(Fig. 1).

If patients with these responses are subjected to placebo blocks[16], it transpires that of those with concordant responses to comparative blocks, 81% do not respond to saline and can be regarded as true-positive responders and 19% do respond to saline; comparative blocks can be reinterpreted as having been false-positive. Of those patients with discordant responses, 65% do not respond to saline and, therefore, despite their paradoxical responses to comparative blocks, they can be interpreted as having true-positive responses; 35% respond to saline and can be confirmed as negative responders. Those with negative responses to comparative blocks remain classified as negative, regardless of their response to placebo blocks (seeFig. 1). Patients with negative responses do not proceed to treatment; only those with true-positive responses proceed to treatment. Percutaneous radiofrequency neurotomy offers a success rate for complete relief of pain of about 70%[7].

According to model A, if 1,000 patients respond to an initial diagnostic block, all 1,000 proceed to comparative blocks, where 20 have a negative response, 220 have a discordant response, and 760 have a concordant response(Fig. 1). All 1,000 patients also undergo placebo blocks. Of those who exhibit discordant responses, 143 are found to be negative to placebo and proceed to treatment. After treatment, 100 from the discordant group and 431 from the concordant group obtain relief of their pain. The total costs involved are those of 1,000 initial blocks and 1,000 comparative blocks, plus 1,000 placebo blocks and 759 radiofrequency neurotomies. This model provides 531 successful outcomes and 228 failures.

Figure 2 illustrates model B where placebo blocks are omitted and only comparative blocks are used. The difference in this model is that a diagnostic decision is made on the basis of comparative blocks and, therefore, the physician does not know which patients would have responded to saline blocks under model A, and who, as a result, would not have proceeded to treatment. Consequently, under model B those patients proceed to treatment but are destined to fail on the grounds that they do not actually have zygapophysial joint pain, in which case denervating their joint would have no lasting effect on their pain.

Figure 2

Model B. Flow chart of the process and outcomes of comparative local anaesthetic blocks without placebo controls for cervical zygapophysial joint pain. Note that the number of successes are the same as in model A. The amount that fail in each arm of the flow chart are those in model A plus the amount that, in model A, were found to be positive to placebo.

Numbers in model B are obtained by eliminating the placebo node in each arm of model A and collapsing the treatment nodes toward results of comparative blocks. As a result, the number of patients destined for success is the same as for model A, but the number who fail includes those who would have been positive to placebo blocks. If 1,000 patients enter the process, all 1,000 undergo comparative blocks and 980 proceed to treatment. The total cost of model B is the cost of 1,000 initial blocks, 1,000 comparative blocks, and 980 treatments. This model provides 531 successful outcomes and 449 failures.

Figure 3 illustrates model C, in which control blocks are omitted completely and all patients with a positive response to a single, initial diagnostic block proceed to treatment. The cost of this model is the cost of 1,000 initial blocks and 1,000 treatments. After treatment, 531 patients will be relieved of their pain, as in models A and B, but 469 will fail because negative responders and responders to placebo will not have been identified.

Figure 3

Model C. Flow chart of the process and outcomes of proceeding to treatment without control diagnostic blocks for cervical zygapophysial joint pain. The amount that succeed with treatment is the same as the total amount of successes in model A. The amount that fail is the residual from the amount treated.

With respect to the diagnosis and treatment of lumbar zygapophysial joint pain, different clinical data apply. Lumbar zygapophysial joint pain is less common among patients with back pain[1,2] than cervical zygapophysial joint pain is among patients with neck pain[3,4]. Furthermore, the false-positive rate of lumbar zygapophysial joint blocks is high[11]. Conversely, there are no data on the proportions of patients undergoing lumbar zygapophysial joint blocks who exhibit negative, discordant, and concordant responses and their subsequent responses to placebo blocks. Consequently, the models for the investigation and treatment of lumbar zygapophysial joint pain are relatively simple compared with those available for cervical zygapophysial joint pain.

Figure 4 illustrates model D, in which comparative blocks are used before proceeding to treatment for lumbar zygapophysial joint pain. Research data show that of patients positive to an initial block, only 31% will be positive to a subsequent control block[1]; only those patients would proceed to treatment. In stringently selected patients, the success rate for lumbar radiofrequency neurotomy is close to 90%[6]. Under those conditions, if 1,000 patients are found to be positive to initial blocks, then all 1,000 will proceed to control blocks. Of those, only 310 will have the diagnosis confirmed and only those 310 patients will proceed to treatment, with 279 obtaining a successful result(Fig. 4). The costs of this model are the cost of 1,000 initial blocks, 1,000 control blocks, and 310 treatments. It produces 279 successful outcomes and only 31 failures.

Figure 4

Model D. Flow chart of the process and outcomes of comparative, local anaesthetic diagnostic blocks for lumbar zygapophysial joint pain.

Figure 5 illustrates model E, in which no control blocks are used, and all patients found positive upon an initial block proceed to treatment. The costs of this model are the cost of 1,000 blocks and 1,000 treatments. It produces 279 successful outcomes and 721 failures.

Figure 5

Model E. Flow chart of the process and outcomes of proceeding to treatment of cervical zygapophysial joint pain without control blocks. The amount that succeed with therapy is the same as in model D. The amount that fail is the residual from the amount treated.

For the treatment of cervical zygapophysial joint pain, models A,B, and C each produce the same number of successful outcomes after treatment. Their relative cost-effectiveness, therefore, is only a function of their relative costs. The models can be compared algebraically to determine which is the less expensive under certain conditions(Table 1).

View this table:
Table 1

The relative costs of models A, B, and C for the diagnosis and treatment of cervical zygapophysial joint pain for 1,000 patients

ModelDescriptionCost
Cervical
APlacebo-controlled blocks759 CT + 3,000 CB
BComparative blocks only980 CT + 2,000 CB
CNo control blocks1,000 CT + 1,000 CB
Lumbar
DComparative blocks310 CT + 2,000 CB
ENo control blocks1,000 CT + 1,000 CB
  • CB, cost of block; CT, cost of treatment.

Let CT = k.CB, where CT is the cost of treatment and CB is the cost of blocks. From the data inTable 1:

Cos of model A = 759(k.CB) + 3000CB

Cos of model B = 980(k.CB) + 2000CB

Cos of model C = 1000(k.CB) + 1000CB

Cos of model D = 310(k.CB) + 2000CB

Cos of model E = 1000(k.CB) + 1000CB

These equations indicate how the cost-effectiveness of different regimens of diagnostic blocks is a function of the relative costs of blocks and treatment. Model B (comparative blocks) is more cost-effective than model C (no controls) whenever k is greater than 50 (i.e., when the cost of treatment is more than 50 times the cost of a block). When k is less than 50 (i.e., when the cost of treatment is less than 50 times the cost of a block), using no controls becomes more cost-effective than comparative blocks. It is perhaps this difference that has caused some individuals intuitively to consider proceeding directly to treatment to be less expensive on comparative blocks.

Model A (placebo blocks) is more cost-effective than model C (no control blocks) whenever k is greater than 8.3 (i.e., when the cost of treatment is more than 8.3 times the cost of a block). Using no controls become more cost-effective than placebo blocks only when k is less than 8.3. For model B (comparative blocks) to be more cost-saving than model A (placebo blocks), k must be less than 4.53 (i.e., the cost of treatment must be less than 4.53 times the cost of a block). If k is greater than 4.53, placebo blocks become more cost-effective than comparative blocks.

With respect to lumbar zygapophysial joint pain, the comparisons are less convoluted. For proceeding directly to treatment to be more cost effective than relying on comparative blocks, k must be less than 1.45. Thus, although the costs of treatment are greater than 1.45 times the costs of a block, avoiding control blocks is not justified. The critical value of k for lumbar zygapophysial joint pain is less than that for cervical zygapophysial joint pain because of the lesser prevalence of lumbar zygapophysial joint pain.

Discussion

The costs of diagnostic blocks and of radiofrequency neurotomy differ from country to country and from provider to provider. Therefore, there is no single solution to the equations developed in this study. However, they are of a form that allows any user to substitute their own, local, or preferred values for the unknown variables. The critical factor is the value of k, which is the ratio between the cost of treatment and the cost of blocks, and which is independent of differences in absolute costs between countries or between providers.

One set of possible solutions is provided inTable 2. This lists the reimbursements for diagnostic blocks and radiofrequency neurotomy in Australia, under Medicare and under the schedule of fees recommended by the AMA; and from two sources in the United States, using various schedules.

View this table:
Table 2

The comparative costs of treatment of zygapophysial joint pain by radiofrequency neurotomy, and its diagnosis by medial branch blocks, based on fees applied in Australia and in the USA

Cost of treatmentCost of block
Source and itemCodeItem cost ($)Total cost ($)Item cost ($)Total coste ($)k
Australia
  Medicare rates419686.16
    For two nerves26849
    Anaesthesia49
    Assistant39
    Fluoroscopy6319
  AMA rates1,0561328.00
    For two nerves57097
    Anaesthesia49
    Assistant83
    Fluoroscopy11435
    Day admission*240
USA
  1a1,2277931.55
    For two nerves235173
    Facility fee992620
  1b1,3109551.37
    For two nerves318335
    Facility fee992620
1c1,2969251.40
For two nerves304305
Facility fee992620
  2a2481731.43
    For two nervesCPT64442.2215281
    SurchargeCPT6443.225450
    FluoroscopyCPT7600.264242
    ConsumablesHCPC A455000
  2b4712841.66
    For two nervesCPT64442.22222155
    SurchargeCPT6443.2216747
    FluoroscopyCPT7600.265757
    ConsumablesHCPC A45502525
  2c4994001.25
    For two nervesCPT64442.22307200
    SurchargeCPT6443.2211096
    FluoroscopyCPT7600.265757
    ConsumablesHCPC A45502547
  • * In Australia, a facility fee for a day-stay admission is chargeable only for privately insured patients; a corresponding item is not chargeable for patients treated under Medicare, and is therefore not included in the costs shown in the table under Medicare. The US fees are derived from two states (1 and 2), each with three payers (a, b, c). Australian costs have been normalized to US dollars, at an exchange rate of $ AUS 1.00 = $US 0.60.

  • k is the ratio between the cost of treatment and the cost of blocks.

The data inTable 2 show that, for cervical zygapophysial joint blocks, neither of the systems of reimbursement in Australia reward best practice; k does not exceed the critical value of 8.3, making placebo-controlled blocks more cost-effective than no-controlled blocks. However, the AMA rates fall just short of the critical value. In both instances, however, because k is greater than 4.53, placebo-controlled blocks are more cost-effective than comparative blocks. With regard to lumbar zygapophysial joint bocks, because k is greater than 1.45, controlled blocks are cost-effective under both Medicare and the AMA schedule.

Table 3 lists the actual, computed costs. It shows that under Medicare for cervical zygapophysial joint pain, placebo controls are more expensive than no blocks by about $35 per patient, on average; under AMA rates, placebo controls are more expensive than no controls, but only by about $9.50 per patient. For lumbar zygapophysial joint pain, model E (no controls) is nearly twice as costly as model D (comparative blocks).

View this table:
Table 3

Total costs of diagnosing and treating 1,000 patients with cervical or lumbar zygapophysial joint pain

Model
CervicalLumbar
SystemA Placebo controlsB Comparative blocksC No controlsD Comparative blocksE No controls
Australia
  Medicare  522,021  546,620  487,000  265,890  487,000
  AMA1,197,5041,298,8601,188,000  591,3601,188,000
USA
  1a3,310,2932,788,4602,020,0001,966,3702,020,000
  1b3,859,2903,193,8002,265,0002,316,1002,265,000
  1c3,758,6643,120,0802,221,0002,251,7602,221,000
  2a  707,232  589,040  421,000  422,880  421,000
  2b1,209,4891,029,580  755,000  714,010  755,000
  2c1,578,7411,289,020  899,000  954,690  899,000
  • Total costs are derived by substituting the costs recorded inTable 2 into the summary costs for each model recorded inTable 1. Australian costs have been normalized to US dollars, at an exchange rate of $AUS 1.00 = $US 0.60.

The situation is different in the United States. The available figures indicate that the reimbursement for radiofrequency neurotomy is barely greater than that for blocks, and both are dwarfed by the facility fee when this applies(Table 2). As a result, k ranges in value from 1.25 to 1.66. From a financial perspective, this renders any controlled block cost-ineffective for cervical zygapophysial joint pain. Yet for lumbar zygapophysial joint pain, comparative blocks are, in fact, more cost-effective than no controls, although only under some schedules of reimbursement. For cervical zygapophysial joint pain, the actual total costs are consistently, and often substantially, greater for any form of controlled block than for no controls(Table 3). For lumbar zygapophysial joint pain, however, controlled blocks are sometimes patently cheaper, or only marginally more expensive than no controls(Table 3).

The travesty that arises is that the fee schedule in the United States prohibits best practice, as does the Medicare system in Australia. Individuals wishing to perform controlled blocks are prevented or discouraged from doing so for purely financial reasons. Absurdly, it is cheaper to offer speculative treatment and suffer inordinate failure rates than to make an accurate diagnosis and optimize success rates.

The basis for this dysfunctional situation is partly the small differential, in the United States compared with Australia, between the fee for blocks and the fee for radiofrequency neurotomy. In Australia, radiofrequency neurotomy is regarded, classified, and reimbursed as a neurosurgical procedure. In the United States it is reimbursed as little more than another block. However, the greater source of absurdity is the burden of the facility fee. As long as the facility fee is 3 times the cost of either radiofrequency neurotomy or a diagnostic block, it will never be more cost-effective to perform controlled blocks. As a result, American patients are forced by economic pressures not to obtain an accurate diagnosis and to suffer the capriciousness of presumptive therapy. The ethics and practicalities of this method need to be considered.

Model C (no controlled blocks) achieves the same number of successes as model A ( placebo blocks), but incurs 469 failures for every 531 successes. Model A incurs only 228 failures. Insurers and doctors wishing to sustain model C must realize that they are paying for predictable failures. Patients are entitled to be informed that despite having to endure a long procedure that involves radiation exposure and the risk of complications, they have little more than a 50% chance of benefit from treatment under model C.

The figures are strikingly different in the case of lumbar zygapophysial joint pain. Regardless of the apparent cost-effectiveness, model E (no controlled blocks) will incur 721 failures with treatment for every 279 successes, whereas model D (comparative blocks) will incur only 31 failures. What insurers are funding under these conditions, in the name of cost-effectiveness, are an avoidable 690 failures for every 1,000 patients treated. This is inefficient at best and unethical at worst.

Incalculable are the costs of the disappointment of patients who fail therapy and the disaffection they develop with the system that subjects them to predictable failure. Moreover, insurers should realize that failures do not avert ongoing costs. Having funded a trial of presumptive therapy, insurers are left with 69% of their clients still requiring proper diagnosis and treatment. The opportunity cost is ironic: in the name of apparent cost-effectiveness, surgeons waste their time performing 690 procedures destined to fail, when this time could be spent more productively operating on patients more likely to succeed, if only the diagnosis was made in a more valid way.

These considerations apply only to the immediate, medical costs surrounding zygapophysial joint blocks and radiofrequency neurotomy. The equations provided, however, can be used for broader purposes. The coefficient, k, is the ratio between the costs of treatment and the costs of blocks. These costs, however, could be more than the direct medical costs and could include the costs of accommodation and travel to attend for diagnosis or treatment. Individuals interested in the cost-effectiveness of blocks could determine the value of k if they add their own values of such indirect costs to the costs of blocks and treatment.

Furthermore, the costs used to calculate k could include liability for insurance claims. Diagnostic blocks do not serve simply as an indication for radiofrequency neurotomy. They also provide information that can be critical to proving that a patient has a legitimate source of pain. In that instance, the cost of treatment can become the cost of a settlement. Practitioners who perform diagnostic blocks may be called to testify whether a patient has a genuine (and treatable) source of pain. A practitioner who performs uncontrolled diagnostic blocks cannot testify honestly and accurately; the false-positive rate of single blocks is too high. For medicolegal purposes, only controlled blocks provide valid information and placebo-controlled blocks provide an irrefutable diagnosis. Thus, for medicolegal purposes, uncontrolled blocks are useless; only controlled blocks are useful.

In that context, for controlled blocks to be cost-effective, the cost of a claim needs to be more than 8.22 times the cost of a block in the case of cervical zygapophysial joint pain, or 1.45 times the cost of a block in the case of lumbar zygapophysial joint pain.

In this way, individuals interested in comparing costs, for whatever purpose, can use the equations developed in this study. All that is required is to substitute the values of interest for the unknown variables in the equations provided.

Footnotes

  • This study was supported by a grant from the Motor Accidents Authority of New South Wales and by the National Musculoskeletal Medicine Initiative of the Commonwealth of Australia.

References

View Abstract