The purpose of this study will be to determine the efficacy and the prognostic value of a continuous intrathecal prognostic infusion test in an in-hospital setting for selecting patients who would have better long term outcomes for treatment with intrathecal implantable devices. The investigators will compare the primary outcomes [changes in pain intensity score (NRS), patient global impression of change (PGIC)] before and after intrathecal infusion of an admixture of bupivacaine 0.625 mg/ml and fentanyl 1 mcg/ml versus normal saline.

The study will include 36 patients with intractable chronic low back pain in the setting of lumbar post-laminectomy syndrome or vertebral compression fracture who failed conservative management and are considered candidates for IDDS.

Prior to the implant, the patients will undergo an intrathecal prognostic infusion test with an externalized catheter. Baseline NRS pain scores will be assessed and documented on all patients upon admission to the preoperative area. An intrathecal catheter will be placed in the outpatient procedure suite at the appropriate level for target dermatomes. The needle entry point will occur in the upper lumbar spine and catheter tip will be placed in the lower thoracic spine, under local anesthesia with the patient awake and with minimal or no sedation. The intrathecal infusion will be started using an external pump once patient is in the PACU. The research component is to perform the intrathecal test with normal saline (inactive placebo solution) in addition to a test with fentanyl and bupivacaine (active solution). Patients will be randomly assigned to either Group I (continuous infusion of bupivacaine and fentanyl followed by saline) or Group II (continuous infusion of saline followed by bupivacaine and fentanyl).

In PACU, patients will be started on an infusion rate of 0.5 ml/hr and titrated to pain relief greater than 50% of baseline or up to 0.8-1.0 ml/hr within 6-8 hrs after start of the infusion. A clinician blinded to the treatment arm will assess NRS and PGIC on the patients after approximately 12 hours. Assessment will include changes in pain intensity score at rest and upon ambulating or performing maneuvers that normally elicit patient's low back pain. A 4-6-hour washout period will be allotted with infusion of preservative-free normal saline at a rate of 0.2 ml/hr, after which the physician will document a return of the NRS to baseline before switching therapies.

Recruitment: Subjects will be recruited from patients seen at the University Hospitals Pain Medicine clinics by attending physicians. Screening will be done before obtaining consent by an investigator. If patients are deemed appropriate for an intrathecal device they will undergo normal procedures and guidelines in place prior to being considered candidates for an implantable device. No subject will be compensated for participation.

Consent Process: Subjects will be consented by one of the investigators. An explanation in lay terms for the reasons of the study and the proposed prognostic benefits will be used to promote patient understanding. If interested, eligible individuals will be given the opportunity to ask and have all questions addressed before signing the informed consent document. The procedure will occur at their next visit, and continued consent of study participation will be confirmed.

Study Design: This is a randomized, double-blind, placebo controlled cross over study comparing prognostic intrathecal testing with an admixture of bupivacaine and fentanyl versus saline. None of the procedures in this study deviate from usual clinical care that patients receive at UHCMC or nationally. Baseline scores using the numerical rating scale (NRS) for pain (a scale form 0-10, where 0 signifies no pain at all, and 10 the worst possible pain) will be recorded, both in the sitting or supine position (least pain) and with ambulation or standing (worst pain). Patients will be given weight based cefazolin (or vancomycin if indicated) prior to placing the externalized intrathecal catheter. Placement of the percutaneous intrathecal catheter will be done in the operating room with minimal or no sedation in the prone or lateral decubitus position under fluoroscopic guidance. Needle entry will occur in the mid-upper lumbar spine and through the needle an intrathecal catheter will be advanced until its tip is positioned in the posterior intrathecal space in the lower thoracic spine. The needle is then removed and the catheter is secured in place with steri-strips and a clear sterile bio-occlusive dressing will be placed. Patients will then be transferred to the PACU where they will be initiated on one of two solutions that will be prepared for each patient by the investigational pharmacy staff at UHCMC. The solutions will be labeled as "Intrathecal solution 1" and "Intrathecal solution 2" and will be contained in a sterile 50 ml bag. Solution 1 and 2 may contain either:

1. Preservative-free normal saline
2. Fentanyl 1 mcg/ml and bupivacaine 0.625 mg/ml

The content of Intrathecal solution 1 and 2 will be unknown to all investigators and participants in the study with the exception of the investigational pharmacy. The order of the Intrathecal solution (1 or 2) will be determined by pharmacy using a computer generated random sequence allocation. The intrathecal catheter will be attached to a pump delivering solution 1 or 2 at around noon time, in the recovery area on the day the catheter is placed. A bolus of 1cc will be given through the infusion pump at initiation of therapy and the patient will then be started on an infusion rate of 0.5 ml/hr. After 3-4 hours (around 3-4 pm) and similarly around 6-7 pm the rate will be titrated depending on patient's response up to a maximum of 0.8-1.0 ml/hr. If the patient has achieved > 50% pain relief compared to baseline, no up-titration will occur; i.e. the rate will be increased only if the patient has not had 50% or more reduction in baseline pain on the NRS. The intrathecal rate will be kept the same provided the patient had 50% or greater decrease in pain scores or has reached the 1.0 ml/hr rate (whichever comes first). The rate will be unchanged from 6-7 pm until around 6-7 am the next morning when the infusion will be stopped and the patient will be assessed for pain relief. In the morning, the patient will be asked to rate the pain score at rest (in bed or chair) and with ambulation/standing. The pain scores will be recorded and the catheter will be aspirated at the hub to ensure continued cerebrospinal fluid flow and the patient will be started on a solution of preservative-free normal saline at 0.2 ml/hr to keep the catheter patent. After 6 hours, around noon time, the patient will be crossed over to Intrathecal solution 1 or 2, depending on what she/he had the day before, given 1.0 ml of that solution as a bolus and then infusion will be started at 0.5 ml/hr and the same protocol as the day before will be repeated with the patient discharged the next morning. Patients who experience greater than 50% pain relief (relative to baseline) with either intrathecal solution will be offered the implant of a permanent IDDS that will deliver a combination of bupivacaine and low-dose fentanyl. Patients not responding to both solutions with greater than 50% pain relief will be considered to have failed the intrathecal test and would not proceed to implant. The patients will be asked to pick which solution provided better pain relief: solution 1 or solution 2 and responses will be recorded. Additionally, pain scores obtained periodically as part of patients' usual clinical care vital signs and recorded by the nursing staff on the hospital ward will be collected throughout the study. Un-blinding for patients who had a successful intrathecal prognostic infusion test with greater than 50% pain relief will not occur until 12 months have elapsed since the pump implant.

Outcome measures will include:

1. Baseline prior to commencement of the prognostic infusion test: Pain intensity using the Numerical Rating Scale [NRS], patient global impression of change [PGIC], Oswestry disability index [ODI] and painDETECT.
2. At 14-18 hours: Pain intensity in Numerical Rating Scale [NRS], patient global impression of change [PGIC], complications and side effects.
3. Prior to second infusion: Pain intensity in Numerical Rating Scale [NRS], patient global impression of change [PGIC], complications and side effects.
4. At prognostic infusion test completion: Pain intensity in Numerical Rating Scale [NRS], patient global impression of change [PGIC], complications and side effects, Oswestry disability index [ODI] and painDETECT.
5. At 6 and 12 months post-implant for implanted patients Pain intensity in Numerical Rating Scale [NRS], patient global impression of change [PGIC], Oswestry disability index [ODI] and painDETECT.

Study Methodology/Procedures: The study will include 36 patients with intractable chronic low or mid back pain due to failed back surgery syndrome or vertebral fracture who failed conservative management including epidural steroid injection and medical therapy and were referred to our practice for pain management.

Patient will undergo the usual psychological and medical evaluations before the initiation of the prognostic infusion test. Patients who are considered candidates for intrathecal pump implant fulfilling the inclusion/exclusion criteria above and who elect to participate in the study will be randomly assigned to two groups.

Group I tested with continuous infusion of intrathecal bupivacaine 0.625 mg/ml and fentanyl 1 mcg/ml for 14-18 hours followed by a trial with normal saline for another 14-18 hours.

Group II tested with intrathecal normal saline for 14-18 hours followed by intrathecal Bupivacaine 0.625 mg/ml and fentanyl 1 mcg/ml for another 14-18 hours.

Note that drugs will be delivered by the pharmacy to a blinded physician and labeled as Intrathecal solution 1 and Intrathecal solution 2 to be administered sequentially, separated by a 4-6-hr infusion of preservative-free saline.

Outcomes will be assessed and documented on all patients upon admission to the preoperative area. The patients will be taken then to the procedure room and a standardized intrathecal catheter will be placed under fluoroscopic guidance where the tip of the catheter will be placed at the T7-T11 posterior intrathecal interspace. Patients will be discharged to the PACU where they will be started on a rate of 0.5 ml/hr.

Six to eight hours following initiation of the infusion, all the patients will be titrated to 0.8-1.0 ml/hr, provided less < 50% improvement in pain scores occurs. A physician who is blinded to the treatment will assess NRS after approximately 12 hrs (around 6-7 am of the following day). A 4-6 hours washout period will ensue with infusion of preservative-free normal saline at a rate of 0.2 ml/hr after which the physician will document a return of the NRS to baseline before switching therapies or record the value at 6 hrs after infusing normal saline and switch then to solution 2. NRS will be reassessed around 6 am the following morning. Additionally, pain scores documented with usual clinical care vital signs will be captured. All reported adverse events will be recorded.

No pain medications will be prescribed during the admission. If such medications are needed, the patient will be excluded from continuing on with the study and will be recorded as a prognostic-infusion-test failure.

The intrathecal catheter will be aspirated for confirmation of free cerebrospinal fluid flow (about 1 ml) between all solution changes and at the end of the prognostic infusion test.

After the completion of the prognostic infusion test, the catheter will be removed and patients will be discharged home.

Only patients who report >50% reduction from baseline NRS while receiving either intrathecal solution will be considered for intrathecal drug delivery system implant. It is conceivable that some patients may get >50% reduction in pain scores with both the active solution and saline or have better outcome with the saline solution. The patients will be asked to answer a binary question rating preference to solution 1 vs. solution 2. Patients with pain relief greater than 50% will be implanted with an IDDS and will receive an intrathecal solution of fentanyl and bupivacaine. All subjects with >50% pain relief with either or both intrathecal solutions will be implanted. Even if the patient gets >50% pain relief only from the saline solution, or the patient chooses the saline solution over the active solution (when asked if #1 vs. #2 was better relief) each patient will receive active drug after being implanted. All patients will be compared in long-term outcome (secondary outcome measures) at 6 months and 12 months versus the response to the prognostic test solutions.

Unblinding of solution 1 and 2 will not occur until 12 months have elapsed since pump implant. The six and twelve month visits will be coordinated with a pump refill visit.

Data Collection: Randomization will be performed, and baseline data will collect on admission to the preoperative area. A physician or physician assistant will obtain all data.

Baseline data collected will include name, last 4 digits of social security number, age, sex, race, duration of pain, treatment group, average 0-10 low back numerical rating scale (NRS) pain scores over the past week and analgesic medication consumption.

The primary outcome variable will be the change in pain intensity score [NRS] 0-10 numerical rating scale back pain score at the end of the intrathecal prognostic infusion testing period around 6am between Intrathecal solution 1 and Intrathecal solution 2.

Secondary outcome variables will be Oswestry disability score, changes in painDETECT, Patient Global Impression of Change (PGIC) and side effects (medications) and complications (injections). These variables will be recorded at baseline, at the completion of each phase of the prognostic infusion test, and at 6 and 12 months post-implant.

Inclusion Criteria:

• Previous lumbar or thoracic spine surgery or lower thoracic/lumbar vertebral compression fracture
• Intractable pain of trunk (more than limbs)
• Patient who passed psychological evaluations as part of the usual clinical care prior to consideration of IDDS and are stable with current pain condition and medications
• Failed more conservative management.

Exclusion Criteria:

• Untreated coagulopathy or infection.
• Immune compromised state precluding having an implant.
• Allergic reactions to bupivacaine or fentanyl.
• Pregnancy
• Patients using more than 30 mg oral equivalents of morphine daily or who are unable to wean down below that dosage for more than 4 weeks before the prognostic intrathecal infusion test.
• Neurological deficits characterized as weakness in lower extremities with evidence of nerve damage
• Patients with cognitive disorders who would not be able to provide meaningful outcome responses

The drugs will be delivered by the pharmacy to a blinded physician and labeled as Intrathecal solution 1 and Intrathecal solution 2 to be administered sequentially, separated by a 4-6-hr infusion of preservative-free saline. The content of Intrathecal solution 1 and 2 will be unknown to all investigators and participants in the study with the exception of the investigational pharmacy. The order of the Intrathecal solution (1 or 2) will be determined by pharmacy using a computer generated random sequence allocation.