Infusion Pumps

Number: 0161

(Replaces CPB 110, CPB 338)

Policy

1. Implantable Infusion Pumps

   Aetna considers implanted infusion pumps medically necessary durable medical equipment (DME) when all of the following criteria are met:

   • It is medically necessary that the drug be administered by an implanted infusion pump; and
   • The drug is medically necessary for the treatment of members (see medical necessity criteria for various types of infusion pumps below); and
   • The infusion pump has been approved by the FDA for infusion of the particular drug that is to be administered.

   1. Anti-Spasmodic Drugs

      Aetna considers an implantable infusion pump medically necessary when used to intrathecally administer anti-spasmodic drugs (e.g., baclofen) to treat chronic intractable spasticity in persons who have proven unresponsive to less invasive medical therapy as determined by the following criteria: 

      1. Member has failed a six-week trial of non-invasive methods of spasticity control, such as oral anti-spasmodic drugs, either because these methods fail to adequately control the spasticity or produce intolerable side effects; and
      2. Member has a favorable response to a trial intrathecal dosage of the anti-spasmodic drug prior to pump implantation. 

      Intrathecal baclofen (Lioresal) is considered medically necessary for the treatment of intractable spasticity caused by spinal cord disease, spinal cord injury, or multiple sclerosis and for stiff person syndrome.  Baclofen is considered medically necessary for persons who require spasticity to sustain upright posture, balance in locomotion, or increased function.

      Documentation in the member's medical record should indicate that the member's spasticity was unresponsive to other treatment methods and that the oral form of baclofen was ineffective in controlling spasticity or that the member could not tolerate the oral form of the drug.  A trial of oral baclofen is not a required prerequisite to intrathecal baclofen therapy in children ages 12 years old or less due to the increased risk of adverse effects from oral baclofen in this group.

      The medical record should document that the member showed a favorable response to the trial dosage of the baclofen before subsequent dosages are considered medically necessary.  An implanted pump for continuous fusion is considered not medically necessary for members who do not respond to a 100 mcg intrathecal bolus.
      
      Intrathecal baclofen is considered experimental and investigational as a treatment for neuromyotonia (Isaac's syndrome), hydrocephalus, and rheumatoid arthritis.
      
      Intrathecal baclofen, when combined with other agents (clonidine, hydromorphone, morphine, and zinconotide), is considered experimental and investigational because the safety and effectiveness of these combinations has not been established.

   2. Drugs for the Treatment of Chronic Intractable Pain

      A preliminary trial of intraspinal (epidural or intrathecal) administration of opioid drugs (e.g., morphine), ziconotide (Prialt), and/or clonidine is considered medically necessary for persons with of severe chronic intractable pain of malignant or non-malignant origin that is unresponsive to less invasive medical therapy and:

      1. The member's history must indicate that he or she has not responded adequately to non-invasive methods of pain control, such as systemic opioids (including attempts to eliminate physical and behavioral abnormalities which may cause an exaggerated reaction to pain).

      Note: A 1 to 2 day inpatient stay is considered medically  necessary for a preliminary trial of intraspinal opioid drug administration.
      
      An implantable infusion pump is considered medically necessary when used to administer opioid drugs (e.g., morphine), ziconotide (Prialt), and/or clonidine intrathecally or epidurally for treatment of severe chronic intractable pain of malignant or non-malignant origin in persons who meet criteria above and where the following criteria are met:

      1. A preliminary trial of intraspinal opioid drug administration with a temporary intrathecal/epidural catheter has substantiated adequately acceptable pain relief with a 50 percent reduction in pain, the degree of side effects (including effects on the activities of daily living), and acceptance; and
      2. For nonmalignant pain only,a psychological evaluation has been obtained and indicates that the individual is a favorable candidate for permanent intrathecal pump implantation

         Implantable infusion pumps for intrathecal or epidural infusion of opioids, ziconotide, and clonidine are considered experimental and investigational as a treatment for gastroparesis and for all other indications because their effectiveness for indications other than the one listed above has not been established. (Note: Currently, morphine and ziconotide are the only FDA-approved analgesics for long-term intrathecal infusion [Turk et al, 2011]).

   3. Intrahepatic Chemotherapy Infusion for Liver Metastases from Colorectal Cancer

      Implantable infusion pumps are considered medically necessary for administration of intrahepatic chemotherapy (e.g., floxuridine) to members with primary hepatocellular carcinoma and for metastatic colorectal cancer where metastases are limited to the liver.
      
      Aetna considers "one-shot" arterial chemotherapy for persons with liver metastases from colorectal cancer experimental and investigational.

      Note: An average 3 to 5 days inpatient hospitalization is considered medically necessary for intrahepatic chemotherapy. Hospital discharge is dependent on resolution of pain, nausea and vomiting which complicate the procedure. 

   4. Contraindications to Implantable Infusion Pumps

      Implantable infusion pumps are considered not medically necessary for persons with the following contraindications to implantable infusion pumps: 

      1. Members who have an active infection that may increase the risk of the implantable infusion pump; or
      2. Members whose body size is insufficient to support the weight and bulk of the device; or
      3. Members with known allergy or hypersensitivity to the drug being used (e.g., oral baclofen, morphine, etc.); or
      4. Members with other implanted programmable devices where the crosstalk between devices may inadvertently change the prescription.

   5. Experimental and Investigational Uses of Implanted Infusion Pumps

      Implanted infusion pumps are considered experimental and investigational for all other indications, including any of the following:

      1. Implantable infusion pumps for intrahepatic administration of chemotherapy for indications other than noted above, including treatment of hepatic metastases from cancers other than colorectal cancer; or
      2. Implantable pumps for the infusion of heparin for recurrent thromboembolic disease; or
      3. Implantable pumps for the infusion of insulin to treat diabetes; or
      4. Implantable pumps for the infusion of baclofen for chronic neuropathic pain (e.g., complex regional pain syndrome/reflex sympathetic dystrophy).

      See also CPB 0607 - Anesthetic and Antiemetic Infusion Pumps.

2. External Infusion Pumps

   Aetna considers external infusion pumps medically necessary DME for administration of any of the following medications:

   1. Certain parenteral anticancer chemotherapy drugs (e.g., cladribine, fluorouracil, cytarabine, bleomycin, floxuridine, doxorubicin, vincristine, vinblastine, cisplatin, paclitaxel) if the drug is part of an evidence-based chemotherapy regimen and parenteral infusion of the drug at a strictly controlled rate is necessary to avoid systemic toxicity or adverse effects, and the drug is administered either:
      
      

      1. by continuous infusion over 8 hours; or
      2. by intermittent infusions lasting less than 8 hours that do not require the person to return to the physician's office prior to the beginning of each infusion; or
         
         
   2. Certain parenteral antifungal or antiviral drugs (e.g., acyclovir, foscarnet, amphotericin B, or ganciclovir); or
   3. Chemotherapy for primary hepatocellular carcinoma or colorectal cancer where the tumor is unresectable or the member refuses surgical excision of the tumor; or
   4. Deferoxamine for the treatment of acute iron poisoning and iron overload (only external infusion pumps are considered medically necessary); or
   5. Heparin for the treatment of thromboembolic disease and/or pulmonary embolism (only external infusion pumps used in an institutional setting are considered medically necessary); or
   6. Heparin to adequately anticoagulate women throughout pregnancy (warfarin compounds are not routinely used for this indication); or
   7. Insulin for persons with diabetes mellitus who meet the selection criteria for external insulin infusion pumps for diabetes set forth below; or
   8. Morphine or other narcotic analgesics (except meperidine) for intractable pain caused by cancer; or
   9. Other parenterally administered drugs where an infusion pump is necessary to safely administer the drug at home; or
   10. Parenteral epoprostenol or treprostinil for persons with pulmonary hypertension; or
   11. Parenteral inotropic therapy with dobutamine, milrinone, and/or dopamine; or
   12. Post-operative nerve block.

   Aetna considers external infusion pumps experimental and investigational for all other indications.  See also CPB 0468 - Terbutaline Pump for Preterm Labor.

3. External Insulin Infusion Pumps for Diabetes

   Aetna considers external insulin infusion pumps medically necessary DME for the persons with diabetes who meet the criteria in section A or in section B below:

   1. Members must meet all of the following criteria: 

      1. The member has been on a program of multiple daily injections of insulin (i.e., at least 3 injections per day), with frequent self-adjustments of insulin dose for at least 6 months prior to initiation of the insulin pumpFootnotes^*; and 
      2. The member has completed a comprehensive diabetes education program; and
      3. The member has documented frequency of glucose self-testing an average of at least 4 times per day during the 2 months prior to initiation of the insulin pump**; and
      4. The member meets at least one of the following criteria while on multiple daily injections (more than 3 injections per day) of insulin: 
          
         
         1. Dawn phenomenon with fasting blood sugars frequently exceeding 200 mg/dL; or
         2. Elevated glycosylated hemoglobin level (HbA1c greater than 7.0%, where upper range of normal is less than 6.0%; for other HbA1c assays, 1% over upper range of normal); or
         3. History of recurring hypoglycemia (less than 60 mg/dL); or
         4. History of severe glycemic excursions; or
         5. Wide fluctuations in blood glucose before mealtime (e.g., pre-prandial blood glucose levels commonly exceed 140 mg/dL);

      or

   2. The member has been on a pump prior to enrollment in Aetna, and has documented frequency of glucose self-testing an average of at least 4 times per day during the month prior to Aetna enrollment. 

      Aetna considers external infusion pumps for diabetes experimental and investigational where the above-listed criteria are not met.

      Footnote:

      Footnotes^* It may be considered medically necessary to initiate the use of insulin infusion pumps during pregnancy earlier than the criteria stated above to avoid fetal and maternal complications of diabetes and pregnancy. It may be considered medically necessary for poorly controlled women with diabetes to sometimes get started on the pump pre-pregnancy or in the first trimester. 

      Notes on External Insulin Infusion Pumps:

      • Aetna considers a programmable disposable external insulin infusion pump (e.g., OmniPod Insulin Management System) an acceptable alternative to a standard insulin infusion pump for persons who meet medical necessity criteria for external insulin infusion pumps.
      • Aetna's medical necessity criteria for external infusion pumps for diabetes have been adapted from Medicare national policy on external insulin infusion pumps, as outlined in CMS's Coverage Issues Manual Section 60-14.
      • Documentation of continued medical necessity of the external insulin infusion pump requires that the member be seen and evaluated by the treating physician at least once every 6 months.
      • External subcutaneous insulin infusion pumps are only considered medically necessary for persons who have demonstrated ability and commitment to comply with a regimen of pump care, frequent self-monitoring of blood glucose, and careful attention to diet and exercise.
      • Some external insulin infusion pumps (e.g., Paradigm Real-Time Insulin Pump and Continuous Glucose Monitoring System, Animas OneTouch PING, Animas VIBE) are able to take results of the blood glucose reading, calculate the appropriate insulin infusion rate, wirelessly transmit the results from the blood glucose monitor to the pump, and automatically adjust the insulin infusion rate, saving the member some extra steps. These insulin pump features, when present, are considered integral to the external insulin infusion pump and blood glucose monitor.
      • The pump must be ordered by and follow-up care of the member must be managed by a physician with experience managing persons with insulin infusion pumps and who works closely with a team including nurses, diabetic educators, and dieticians who are knowledgeable in the use of insulin infusion pumps.

      See also CPB 0070 - Diabetes Tests, Programs and Supplies.

4. Supplies and Drugs used with Implantable or External Infusion Pumps

   Aetna considers supplies that are needed for the effective use of the DME medically necessary. 

   Such supplies include those drugs and biologicals that must be put directly into the equipment in order to achieve the therapeutic benefit of the DME or to assure the proper functioning of the equipment.

5. Replacement Pumps

   1. Replacement of a functioning insulin pump with an insulin pump with wireless communication to a glucose monitor is considered not medically necessary as such wireless communication has not been shown to improve clinical outcomes
   2. Replacement of an external insulin pump is considered medically necessary for children who require a larger insulin reservoir.
   3. The replacement of infusion pumps that are out of warranty, are malfunctioning, and cannot be refurbished is considered medically necessary.
       

6. Nonprogrammable Disposable Insulin Delivery Systems

   Aetna considers nonprogrammable disposable insulin delivery systems (e.g., V-Go™ disposable insulin delivery device) experimental and investigational because their effectiveness has not been established. 

Background

Implantable infusion pumps are subcutaneously inserted devices that deliver drugs through central venous, intra-arterial, intrathecal or intraperitoneal catheters. Implantable infusion pumps allow drug delivery directly to specific sites and can be programmed for continuous or variable rates of infusion. Examples of implantable infusion pumps include Codman 3000, InfusAid Pump. MedStream Programmable Infusion System, Prometra Programmable Infusion Pump System, SynchroMed Infusion System, and SynchroMed II.

Baclofen (Lioresal) is a derivative of gamma aminobutyric acid (GABA) that acts specifically at the spinal end of the upper motor neurons to cause muscle relaxation.  Intrathecal baclofen may be indicated for patients with severe chronic spasticity of spinal cord origin. An implantable infusion pump is required for the administration of intrathecal baclofen.  Intrathecal baclofen therapy is indicated for persons with severe chronic spasticity of spinal cord origin (including multiple sclerosis) that is refractory to oral baclofen or where there are unacceptable side effects from oral baclofen at the effective dose. The patient should be shown to respond to a single intrathecal bolus dose of up to 100 mcgs of baclofen. A positive response is defined as an average two-point drop on an objective muscle tone or spasm screening system (e.g., The Ashworth and Spasm scale). According to available guidelines, intrathecal baclofen therapy is not considered appropriate if the patient has a history of hypersensitivity to Lioresal, is pregnant or has inadequate birth control, has severely impaired renal function, has severe hepatic or gastrointestinal disease, or has cerebral lesions as the source of spasticity.

Brennan and Whittle (2008) stated that continuous infusion of intrathecal baclofen (ITB) via a subcutaneously implanted pump has developed over the past 2 decades as a powerful tool in the management of spasticity in various adult and pediatric neurological conditions. Acting more focally on spinal GABA receptors, ITB causes fewer systemic side effects than orally administered baclofen. The result is facilitation of daily caring, and symptomatic relief from painful spasm. With increasing experience of ITB use, novel applications and indications are emerging. These include the management of dystonia and chronic neuropathic pain. However, despite some recent authoritative reviews, there is still uncertainty about optimal use and evaluation of this therapy.

Shilt et al (2008) stated that ITB is an effective treatment of spasticity in patients with cerebral palsy (CP). However, several recent reports have raised concerns that the treatment may be associated with a rapid progression of scoliosis. The objective of this study was to further examine the effect of ITB treatment on the progression of scoliosis in patients with CP. Spastic CP patients who were ITB candidates were followed radiographically. Baseline Cobb angles of the primary curve were measured during the period of ITB pump insertion and at the most recent follow-up visit. Each patient was matched with a control patient by the diagnosis of CP, age, sex, topographical involvement, and initial Cobb angle. The mean rate of change in Cobb angle was compared between ITB and control patients using paired-t test. A multiple linear regression model was used to examine the difference, controlling for age, sex, topographical involvement, and initial Cobb angle. A total of 50 ITB patients and 50 controls were included in the analysis. There was no statistically significant difference between the mean change in Cobb angle in ITB patients (6.6 degrees per year) compared with the matched control patients (5.0 degrees per year, p = 0.39). The results from the multiple regression analysis also failed to show a statistically significant difference (0.92 degrees per year difference between ITB patients and controls, p = 0.56). The authors concluded that the progression of scoliosis in CP patients with ITB treatment is not significantly different from those without ITB treatment. The findings suggest that patients receiving ITB experience a natural progression of scoliosis similar to the natural history reported in the literature.

The discovery of spinal opiate receptors, and that the binding of morphine at relatively low concentrations to these receptors produced effective analgesia have led to the development of intraspinal analgesia for the management of pain.  This mode of opioid administration has become an attractive alternative for cancer patients whose pain is not relieved by conventional drugs and/or routes; and for others who cannot tolerate the side effects of systemic administration of opioids in the dose needed for adequate analgesia as the disease progresses.  A popular procedure for intraspinal administration of opioids analgesics is the implantation of an infusion pump which allows the direct delivery of opioids to the receptors in the spinal cord continuously and/or in an intermittent manner.

Available evidence indicates that chronic intrathecal opioids administration via implantable pumps can provide satisfactory pain relief for patients who suffer from intractable cancer pain.  In addition, it allows the patients to be less dependent on hospital services, thus improving the quality of their lives. Studies have also shown that the same method of treatment was successful in providing quality analgesia to carefully selected patients who experienced chronic pain from nononcologic origins, although reductions in pain and improvements in function is observed less consistently in noncancer pain. Some investigators have reported untoward side effects of intrathecal opioid administration, including development of a fibrous mass around the tip of the catheter, resulting in compression of the intrathecal space with displacement of the spinal cord.

To be considered for spinal analgesia, a patient must have a normal platelet count and no coagulation disorder, infection, or other problems that might preclude the use of spinal drugs.  Before the implantation of a permanent infusion system, an efficacy test is usually performed to assess the patient's response and dose.  One or several trial doses of 5 to 10 mg of epidural morphine, or 0.5 to 1.0 mg intrathecal morphine is/are administered while all other analgesic medications are stopped.  Subjective pain ratings and undesirable side effects are evaluated for several days.  A decision to implant the pump is made only if pain is markedly reduced and other opioid analgesics are not needed by the patient.

University of California-San Francisco’s webpage on “Intrathecal Drug Delivery” (2015) states that “Before a permanent pain pump is implanted, most patients have a trial procedure to determine if they are a good candidate for a permanent pump.  This usually involves using a needle to place a temporary spinal catheter, followed by an inpatient stay of several days while pain medication is infused through it.  In other cases, a single or series of pain medication injections around the spine are given on an outpatient basis”.

TouchStone Interventional Pain Center states that “Although pump placement is an inpatient procedure, most patients spend only 1 night in the hospital for observation”.

Furthermore, Wilkes (2014) stated that “The Polyanalgesic Consensus Conference (PACC) guidelines strongly recommend at least 24-hour inpatient observation for trialing”.

Insulin Pumps

Insulin pumps are devices used to deliver insulin in a programmed and controlled manner to diabetic individuals. These devices work with a separate glucometer through manual or remote functions. The goals of insulin pump therapy are to achieve near-normal control of blood glucose levels. Insulin pumps are categorized as follows:

External insulin pumps deliver insulin via subcutaneous or intraperitoneal routes. External insulin pumps may be either disposable or have disposable components. The following are examples of FDA cleared external insulin pumps.

ACCU-CHEK Spirit Insulin Pump: Non disposable insulin pump with programmable reminders for basal and bolus delivery of insulin. Bluetooth capable for 2-way wireless communication with meter to deliver insulin and adjust pump settings remotely.

OmniPod System is a disposable insulin pump that consolidates the pump, tubing and subcutaneous needle into one compact unit. The unit is worn up to three days before requiring replacement.

One-Touch Ping is a non-disposable insulin pump with programmable basal and bolus insulin delivery that uses a meter with remote capability.

t:slim and t:flex Insulin Delivery System is a battery operated infusion pump capable of both basal and bolus delivery of insulin. It utilizes a motor-driven mechanism to deliver insulin from a disposable cartridge. Unlike other pumps which deliver the insulin by applying pressure from behind the full contents of the syringe/reservoir, this pump withdraws small amounts of insulin from its reservoir and transfers it into the micro-delivery chamber before being delivered into the subcutaneous tissue at the correct increment and rate.

V-Go Disposable Insulin Delivery Device is a fully disposable, non-electronic device for the delivery of basal-bolus insulin therapy for adults with diabetes. It provides a continuous preset basal rate of insulin and allows for on-demand bolus dosing around mealtimes. It is a delivery option for diabetics and is applied to the skin daily for one 24-hour period.

Implantable insulin pumps deliver insulin via intraperitoneal or intravenous routes. Currently there are not any devices that have received FDA approval for use outside of a clinical trial.

Aetna's medical necessity criteria for external infusion pumps for diabetes have been adapted from Medicare national policy on external insulin infusion pumps, as outlined in CMS's Coverage Issues Manual Section 60-14.

There is some limited evidence that external insulin infusion pumps improve glycemic control over multiple daily injections in persons with type 2 diabetes (Hammond, 2004; Pickup & Renard, 2008; Fatourechi, et al., 2009), although not all studies have been consistent (see, e.g., Raskin, et al., 2003; Herman, et al., 2005; Berthe, et al., 2007; Parkner, et al., 2008).

Fatourechi et al (2009) reported on a metaanalysis of randomized controlled clinical trials of continuous subcutaneous insulin infusion (CSII) over multiple daily injections (MDI) in persons with diabetes.  The investigators identified 15 eligible randomized trials of moderate quality, with elevated baseline and end-of-study hemoglobin A1c (HbA1c) levels.  The investigators reported that patients with type 1 diabetes using CSII had slightly lower HbA1c [random-effects weighted mean difference, -0.2 %; 95 % confidence interval (CI): -0.3 to -0.1, compared with MDI], with no significant difference in severe (pooled odds ratio, 0.48; 95 % CI: 0.23 to 1.00) or nocturnal hypoglycemia (pooled odds ratio 0.82, 95 % CI: 0.33 to 2.03).  Adolescents and adults with type 1 diabetes enrolled in cross-over trials had no-nsignificantly fewer minor hypoglycemia episodes per patient per week (-0.08; 95 % CI: -0.21 to 0.06) with CSII than MDI; children enrolled in parallel trials had significantly more episodes (0.68; 95 % CI: 0.16 to 1.20; p (interaction) = 0.03).  The investigators reported that outcomes were not different in patients with type 2 diabetes.  The investigators concluded that "[c]ontemporary evidence indicates that compared to MDI, CSII slightly reduced HbA1c in adults with type 1 diabetes, with unclear impact on hypoglycemia.  In type 2 diabetes, CSII and MDI had similar outcomes.  The effect in patients with hypoglycemia unawareness or recurrent severe hypoglycemia remains unclear because of lack of data."

Guidelines from the American Association of Clinical Endocrinologists (Robard, et al., 2007) provide a grade C recommendation for patients with type 2 diabetes, "[c]onsider use of continuous subcutaneous insulin infusion in insulin-treated patients." Guidelines from the National Institute for Health and Clinical Excellence (2008) do not recommend use of CSII in persons with type 2 diabetes.

1. C-peptide positive but with suboptimal control on a maximal program of basal/bolus injections;
2. substantial “dawn phenomenon”;
3. erratic lifestyle (eg, frequent long distance travel, shift-work, unpredictable schedules leading to difficulty maintaining timing of meals); and
4. severe insulin resistance, candidate for U500 insulin by continuous subcutaneous insulin infusion (CSII). 

The consensus statement said that current literature on insulin pump use has focused primarily on the benefits of CSII in patients with type 1 diabetes mellitus, with some attention to the role of CSII in patients with severely insulin-deficient type 2 diabetes mellitus. The consensus statement indicated that "few clinical investigations have examined CSII use in patients with type 2 DM." The consensus statement noted that one analysis of 4 randomized controlled trials in patients with type 2 diabetes mellitus (citing Monami, et al.) found no significant HbA1c improvements and no significant differences in hypoglycemic risk with CSII versus multiple daily injection therapy over 12 weeks. The consensus statement noted, however, that there was a nonsignificant trend toward decreased insulin requirements was observed among CSII patients. The statement also cited other recent meta-analysis of insulin pump therapy in type 2 diabetes, one by Jeitler, et al. (2008) which they interpreted as finding "no conclusive CSII benefits for patients with type 2 DM", and a meta-analysis by Fatourechi, et al. (2009), which they interpreted as finding "CSII and MDI outcomes were similar among patients with type 2 DM." The consensus statement also referenced a nested case-control study of the use of insulin infusion pumps versus non-pump therapy in pregnant women with type 2 diabetes or gestational diabetes (Simmons, et al., 2001).

A systematic evidence review by Mukhopadhyay et al (2007) found no advantage of using continuous subcutaneous insulin infusion (CSII) over multiple daily injections in pregnant women with diabetes.  The investigators identified randomized controlled clinical trials of the effects of CSII versus multiple daily insulin injections on glycemic control and pregnancy outcome in women with diabetes.  Studies were rated for quality independently by 2 reviewers.  Summary weighted mean differences and odds ratios were estimated for insulin dose, birthweight, gestational age, mode of delivery, hypoglycemic/ketotic episodes, worsening retinopathy, neonatal hypoglycemia, and rates of intrauterine fetal death.  Six randomized clinical trials met the inclusion criteria.  The investigators found that pregnancy outcomes and glycemic control were not significantly different among treatment groups.  The investigators found higher numbers of ketoacidotic episodes and diabetic retinopathy in the CSII group, but these differences did not reach statistical significance.  The investigators found that this systematic review did not find any advantage or disadvantage of using CSII over multiple daily injections in pregnant diabetic women.

1. the submission from the pump users group -- Insulin Pump Therapy (INPUT);
2. interviews with parents of young children who were members of INPUT;
3. some recent studies; and
4. from a summary of findings from the previous assessment report. 

Economic modeling used the Center for Outcomes Research (CORE) model, through an arrangement with the NICE and the pump manufacturers, whose submission also used the CORE model. The 74 studies used for analysis included 8 randomized controlled trials (RCTs) of CSII versus analog-based MDI in either T1DM or T2DM, 8 new (since the last NICE appraisal) RCTs of CSII versus NPH-based MDI in T1DM, 48 observational studies of CSII, 6 studies of CSII in pregnancy, and 4 systematic reviews. The following benefits of CSII were highlighted: better control of blood glucose levels, as reflected by HbA1c levels, with the size of improvement depending on the level before starting CSII; reduction in swings in blood glucose levels, and in problems due to the dawn phenomenon; fewer problems with hypoglycemic episodes; reduction in insulin dose per day, thereby partly off-setting the cost of CSII; improved quality of life, including a reduction in the chronic fear of severe hypoglycemia; more flexibility of lifestyle -- no need to eat at fixed intervals, more freedom of lifestyle and easier participation in social and physical activity; and benefits for the patients' family. The submission from INPUT emphasised the quality of life gains from CSII, as well as improved control and fewer hypoglycemic episodes. Also, there was a marked discrepancy between the improvement in social quality of life reported by successful pump users, and the lack of convincing health-related quality of life gains reported in the trials. With regard to economic evaluation, the main cost of CSII is for consumables, such as tubing and cannulae, and is about 1800 to 2000 pounds per year. The cost of the pump, assuming 4-year life, adds another 430 to 720 pounds per year. The extra cost compared with analog-based MDI averages 1700 pounds. Most studies, assuming a reduction in HbA1c level of 1.2 %, found CSII to be cost-effective. The authors concluded that based on the totality of evidence, using observational studies to supplement the limited data from RCTs against best MDI, CSII provides some advantages over MDI in T1DM for both children and adults. However, there was no evidence that CSII is better than analog-based MDI in T2DM or in pregnancy. They stated that further trials with larger numbers and longer durations comparing CSII and optimized MDI in adults, adolescents and children are needed. In addition, there should be a trial of CSII versus MDI with similar provision of structured education in both arms. A trial is also needed for pregnant women with pre-existing diabetes, to investigate using CSII to the best effect.

Zu et al (2011) described the preparation and characterization of chitosan (CS)-polyvinyl alcohol (PVA) blend hydrogels for the controlled release of nano-insulin; CS- PVA blend hydrogels were prepared using glutaraldehyde as the cross-linking agent.  The obtained hydrogels, which have the advantages of both PVA and CS, can be used as a material for the transdermal drug delivery (TDD) of insulin.  The nano-insulin-loaded hydrogels were prepared under the following conditions: 1.2g of polyethylene glycol, 1.5 g of CS, 1.2 g of PVA, 1.2 ml of 1 % glutaraldehyde solution, 16 ml of water, and 40 mg of nano-insulin with 12 mins of mixing time and 3 mins of cross-linking time. The nano-insulin-loaded hydrogels were characterized using scanning electron microscopy, energy dispersive spectrometry, Fourier-transform infrared spectroscopy, differential scanning calorimetry, thermo-gravimetric analysis, X-ray diffraction, and its mechanical properties were analyzed.  The results showed that all molecules in the hydrogel have good compatibility and they formed a honeycomb-like structure.  The hydrogel also showed good mechanical and thermal properties.  The in-vitro drug release of the hydrogel showed that the nano-insulin accorded with Fick's first law of diffusion and it has a high permeation rate (4.421 μg/(cm(2)hr)).  The authors concluded that these results suggested that the nano-insulin-loaded hydrogels are a promising non-invasive TDD system for diabetes chemotherapy.

Ito et al (2012) developed a dissolving microneedle (DM) application system, where 225 to 300 insulin-loaded DMs were formed on a chip.  After the heat-sealed sheet is removed, the system covered with the press-through package layer is put on the skin.  By pressing with the hand, insulin DMs were inserted into the skin.  Factors affecting the penetration depth of DM were studied using applicator in-vitro and in-vivo experiments.  The penetration depth was determined for rat and human skin.  Two-layered DM array chips were prepared to obtain complete absorption of insulin and administered to the rat abdominal skin.  Plasma glucose levels were measured for 6 hrs.  By comparing the hypoglycemic effect with that obtained after subcutaneous injection, relative pharmacological availability was determined.  The penetration depth increased from 21 +/- 3 μm to 63 +/- 2 μm in proportion to application speed to isolated rat skin, at 0.8 to 2.2 m/s.  Human skin showed similar results in the penetration depth.  The in-vivo penetration depth was dependent on the force (0.5 to 2.5 N) and duration (1 to 10 mins), as the secondary application force.  The penetration depth was 211 +/- 3 μm with 3-min duration in the in-vivo rat experiment.  Dissolving microneedle array chips having an insulin-loaded space of 181.2 +/- 4.2 and 209 +/- 3.9 μm were evaluated in the rat.  Relative pharmacological availability values of insulin from DMs were 98.1 +/- 0.8 % and 98.1 +/- 3.1 %, respectively.  The authors concluded that these findings suggested the usefulness of the 2-layered DM application system for the transdermal delivery of insulin.

Jahn et al (2013) noted that as all major insulin pump manufacturers comply with the international infusion pump standard EN 60601-2-24:1998, there may be a general assumption that all pumps are equal in insulin-delivery accuracy.  These researchers investigated single-dose and averaged-dose accuracy of incremental basal deliveries for 1 patch model and 3 durable models of insulin pumps.  For each pump model, discrete single doses delivered during 0.5 U/hr basal rate infusion over a 20-hr period were measured using a time-stamped micro-gravimetric system.  Dose accuracy was analyzed by comparing single doses and time-averaged doses to specific accuracy thresholds (± 5 % to ± 30 %).  The percentage of single doses delivered outside accuracy thresholds of ± 5 %, ± 10 %, and ± 20 % were as follows: Animas OneTouch® Ping® (43.2 %, 14.3 %, and 1.8 %, respectively), Roche Accu-Chek® Combo (50.6 %, 24.4 %, and 5.5 %), Medtronic Paradigm® Revel™/Veo™ (54.2 %, 26.7 %, and 6.6 %), and Insulet OmniPod® (79.1 %, 60.5 %, and 34.9 %).  For 30 mins, 1 hr, and 2 hrs averaging windows, the percentage of doses delivered outside a ± 15 % accuracy were as follows: OneTouch Ping (1.0 %, 0.4 %, and 0 %, respectively), Accu-Chek Combo (4.2 %, 3.5 %, and 3.1 %), Paradigm Revel/Veo (3.9 %, 3.1 %, and 2.2 %), and OmniPod (33.9 %, 19.9 %, and 10.3 %).  The authors concluded that this technical evaluation demonstrated significant differences in single-dose and averaged-dose accuracy among the insulin pumps tested.  Differences in dose accuracy were most evident between the patch pump model and the group of durable pump models.  Of the pumps studied, the Animas OneTouch Ping demonstrated the best single-dose and averaged-dose accuracy.

V-Go Disposable Insulin Delivery Device:

A non-programmable insulin delivery system using a transdermal microneedle has been advocated as a needle-free alternative to subcutaneous injection conventionally used to treat type 2 diabetes (Johns et al, 2014; Rosenfeld et al, 2012; Kapitza et al, 2008).  The V-Go is a newly developed insulin delivery system.  The push of a button inserts a needle into the patient once daily and remains attached for 24 hours.  The V-Go is designed to release a set basal rate throughout the day, while allowing patients to provide up to 36 units of on-demand bolus insulin with the manual click of 2 buttons.  It is a spring-loaded device filled daily with rapid-acting insulin that runs without the use of batteries or computer software.  However, the effectiveness of a non-programmable transdermal microneedle insulin delivery in the management of diabetic patients has not been established.

Winter et al (2015) reported on a prospective active comparator study to observe the A1C lowering effects of multiple daily insulin injections (MDII) versus the use of the V-Go insulin delivery system for patients with uncontrolled type 2 diabetes mellitus over a 3-month period.  The investigators also assessed the effect on insulin requirement for these patients with secondary comparisons of weight, blood pressure, prevalence of hypoglycemic events, and quality of life before and after 3 months of intensified insulin therapy with regular monitoring by a clinical pharmacist at an internal medicine clinic.  The average A1C lowering experienced by the 3 patients in the V-Go group was 1.5 %, while the average A1C change in the 3 patients in the MDII group was an increase of 0.2 %.  All patients in the V-Go group experienced a decrease in insulin total daily dose (TDD), with an average decrease of 26.3 units.  All patients in the MDII group experienced an increase in insulin TDD with an average of 15 units daily to achieve therapeutic goals individualized for each patient.  All patients who underwent intensification of insulin therapy experienced an increase in subjective quality of life (QOL) as determined using the Diabetes-39 (D-39) questionnaire, though QOL results lacked statistical significance.  Limitations of this study include its observational nature, small size, and limited duration of followup.

Kapitza et al (2008) reported on a proof-of-concept study to evaluate the clinical functionality, safety, and pharmacodynamics of the V-Go delivering insulin aspart and redistributing a single basal dose of insulin glargine as a constant basal infusion supplemented with prandial insulin in subjects with type 2 diabetes mellitus.  In 6 subjects receiving once-daily subcutaneous (SC) injections of insulin glargine (> or = 15 U/day) with or without concomitant oral antidiabetic drugs, glargine was discontinued following a 3-day baseline phase.  The V-Go was then applied to the lower abdomen of the subjects once daily for 7 days (days 1 to 3 inpatient, days 4 to 7 outpatient).  Each V-Go provided a continuous 24-hour preset basal infusion rate of insulin aspart (0.6 U/h) and up to 3 daily prandial doses at mealtimes.  Capillary blood glucose concentrations were measured at 11 time points per day during the baseline and inpatient phases and at 4 time points per day during the outpatient phase.  Additionally, glucose profiles were measured continuously on all days.  The V-Go was well tolerated and operated as anticipated.  The mean +/- SEM prestudy daily dose of SC insulin glargine was 33.3 +/- 13.8 U; the mean daily total insulin aspart dose infused with the V-Go was 31.5 +/- 7.5 and 32.3 +/- 7.8 U for the inpatient and outpatient periods, respectively.  Fasting blood glucose values were similar to those observed at baseline throughout the study, with nonsignificant (NS) reductions in readings collected during the outpatient phase before lunch (-35 +/- 27 mg/dl) and before dinner (-38 +/- 25 mg/dl).  The 2-hour post-prandial glucose trended lower from 231 to 195 mg/dl (NS) at breakfast, 234 to 166 mg/dl (NS) at lunch, and 222 to 171 mg/dl (NS) at dinner.  Bedtime blood glucose decreased (mean change from baseline -52 +/- 21 mg/dl; p = 0.0313), as did nighttime (3:00 AM) measurements (-20 +/- 9 mg/dl; P = 0.0313).  Overall glycemic control tended to improve, as shown by continuous glucose monitoring changing from 173 to 157 mg/dl (p = 0.063, NS) and 156 mg/dl (p = 0.219) during inpatient and outpatient periods, respectively.  Glycemic variability assessed by the M value similarly tended to decrease from 33 +/- 9 to 25 +/- 4 (NS) and 21 +/- 4 (NS) for inpatient and outpatient periods, respectively.  The authors concluded that these first data suggested that use of the V-Go is an attractive alternative to SC insulin injection therapy because metabolic control appears to be maintained or even improved without increasing daily insulin doses.

Rosenfeld et al (2012) described patient perceptions regarding their experience and reported findings in a retrospective analysis of glycemic control in a cohort of patients who used the V-Go, a mechanical, 24-hr disposable, subcutaneous continuous insulin delivery device that delivers a preset basal infusion rate and on-demand insulin.  Patients used the V-Go and answered telephone surveys about their perception of device use.  Corresponding clinical data were retrospectively collected before V-Go initiation, after 12 weeks of use, at the end of treatment, and 12 weeks after discontinuation.  Analyses were performed with non-parametric statistical tests.  A total of 23 patients participated in this study.  Mean values of the following characteristics were documented: patient age, 61 years; body mass index, 30 kg/m2; diabetes duration, 16 years; duration of insulin therapy, 7 years; average duration of V-Go use, 194 days; and mean total daily insulin dose, 50 U at baseline, 46 U while on V-Go, and 51 U after stopping V-Go treatment.  Mean patient rating of the overall experience was 9.1 at 12 weeks on a scale from 1 to 10 (10 being most positive).  Mean hemoglobin A1c value decreased from baseline (8.8 % to 7.6 %; [p = 0.005]) while using the V-Go, and it increased to 8.2 % after treatment.  Fasting plasma glucose trended from 205 mg/dL at baseline to 135 mg/dL while using V-Go and increased to 164 mg/dL after V-Go was stopped.  Weight was essentially unchanged.  No differences in hypoglycemic events were found; site reactions were minor.  The authors concluded that glycemic control improved when patients were switched to the V-Go for insulin delivery, and it deteriorated when the V-Go was discontinued.  The main drawbacks of this study were its retrospective nature, small sample size, and short-term follow-up.  These preliminary findings need to be validated by well-designed studies.

Much of the other data on V-Go are in abstract form.  Furthermore, current guidelines from the American Diabetes Association and the American College of Clinical Endocrinology do not mention the use of a non-programmable disposable insulin delivery systems.

Lajara et al (2016) compared 2 methods of delivering intensified insulin therapy (IIT) in patients with type 2 diabetes inadequately controlled on basal insulin ± concomitant anti-hyperglycemic agents in a real-world clinical setting.  Data for this retrospective study were obtained using electronic medical records from a large multi-center diabetes system.  Records were queried to identify patients transitioned to V-Go disposable insulin delivery device (V-Go) or MDI using an insulin pen to add prandial insulin when A1C was greater than 7 % on basal insulin therapy.  The primary end-point was the difference in A1C change using follow-up A1C results.  A total of 116 patients were evaluated (56 V-Go, 60 MDI).  Both groups experienced significant glycemic improvement from similar mean baselines.  By 27 weeks, A1C least squares mean change from baseline was -1.98 % (-21.6 mmol/mol) with V-Go and -1.34 % (-14.6 mmol/mol) with MDI, for a treatment difference of -0.64 % (-7.0 mmol/mol; p = 0.020).  Patients using V-Go administered less mean ± SD insulin compared to patients using MDI, 56 ± 17 units/day versus 78 ± 40 units/day (p < 0.001), respectively.  Diabetes-related direct pharmacy costs were lower with V-Go, and the cost inferential from baseline per 1 % reduction in A1C was significantly less with V-Go ($118.84 ± $158.55 per patient/month compared to $217.16 ± $251.66 per patient/month with MDI; p = 0.013).  The authors concluded that progression to IIT resulted in significant glycemic improvement.  Insulin delivery with V-Go was associated with a greater reduction in A1C, required less insulin, and proved more cost-effective than administering IIT with MDI.

The main drawbacks of this study included: (i) findings were based on retrospective data, and although careful review was performed to identify patients, a selection bias was possible concerning which intensified insulin regimen was prescribed, (ii) better outcomes in patients prescribed V-Go may be due to a higher level of motivation to control diabetes, as demonstrated by their willingness to try a new method of insulin delivery.  For all patients, diabetes management was provided with no expected frequency of patient contact or mandatory titration, which when enforced may impact A1C reductions.  Prescribed insulin dosing and bolus frequency were based on information obtained from medical records; actual insulin dosing and frequency may have differed across both delivery methods, (iii) patients using MDI were prescribed fewer boluses per day compared to patients using V-Go. Although the ratio of basal to bolus insulin was similar between groups and patients in the MDI group were prescribed significantly more insulin, the difference in bolus frequency may have impacted glycemic control for the MDI group, (iv) insulin adjustments based on meal content and daily glucose profiles were not analyzed due to incomplete data, which limits the ability to fully evaluate daily glycemic control and insulin coverage between groups, and (v) the reporting of hypoglycemic events was based primarily on patient-reported events documented in progress notes and therefore may not accurately depict the risk or severity of hypoglycemia associated with either delivery method.

Combined Intrathecal Baclofen and Other Agents:

Gatscher and associates (2002) stated that complex pain syndromes due to spasticity and central deafferentation often fail to respond to medical therapy and create challenging problems in the pain management.  So far, only spasticity associated musculoskeletal pain has been reported to respond to intrathecal baclofen application.  These investigators reported the treatment of severe neuropathic pain in a patient with erectile dysfunction (ED) and the combined intrathecal application of baclofen and morphine in 5 patients with severe spasticity related pain.  Continuous intrathecal baclofen infusion resulted in a pain-free period of 20 months in the patient with ED.  Patients with spasticity treated with intrathecal application of baclofen and morphine were pain-free for a mean period of 2 years.  The authors concluded that intrathecal baclofen and morphine application was effective in spasticity-related and central deafferentation pain and should therefore be considered in the management of these patients.  The major drawback of this study was its small sample size (n = 5 for patients with spasticity treated with intrathecal baclofen and morphine).

Wallace and colleagues (2010) noted that ziconotide is a non-opioid intrathecal analgesic used to manage moderate to severe chronic pain.  Although ziconotide is approved for intrathecal monotherapy only, the need exists for a critical assessment of the currently available published literature on ziconotide combination therapy.  These investigators evaluated the publications from pre-clinical and clinical peer-reviewed experiments that examined the safety and effectiveness of ziconotide in combination with a variety of other drugs.  A total of 11 relevant publications were identified through a systematic search of multiple databases.  In pre-clinical studies, additive or synergistic anti-nociceptive effects were discovered when ziconotide was used in combination with morphine, clonidine, or baclofen; however, no additional anti-nociceptive effects were observed when bupivacaine was added to ziconotide therapy.  Safety data from animal studies revealed that ziconotide did not exacerbate morphine-induced respiratory depression, or clonidine-induced hypotension or bradycardia; however, ziconotide did potentiate morphine-induced hypotension and inhibition of gastro-intestinal (GI) tract motility.  Results from 2 open-label trials indicated that combination ziconotide and morphine therapy produced greater analgesia than was produced by the use of either drug alone.  Preliminary support for the use of ziconotide in combination with morphine, baclofen, or hydromorphone was provided by case studies.  The authors concluded that although pre-clinical and clinical studies provided some support for the use of ziconotide in combination with morphine, hydromorphone, clonidine, or baclofen, strong evidence-based data are limited.  They stated that controlled, long-term clinical trials are needed.

Saulino (2012) noted that intrathecal therapy has separate indications for refractory pain and spasticity.  Both entities have a relatively high prevalence in neurologic diseases.  This study examined the potential effectiveness of utilizing additive intrathecal morphine (ITM) therapy to a group of patients who had previously stabilized on intrathecal baclofen (ITB) therapy.  Pain intensity was assessed via VASPI; 183 individuals participated in ITB therapy from January 1998 to December 2007; 47 individuals elected to add ITM to their intrathecal therapy regimen; 3 patients were intolerant to ITM/ITB combination therapy.  No significant demographic differences between the 2 groups existed with respect to gender and race.  Non-traumatic and traumatic spinal cord injury (SCI) patients were more likely to participate in combination therapy compared to other diagnoses.  The average stabilized ITB dose for the monotherapy group was not statistically different for the combination therapy group.  The average stabilized ITM for the combination therapy group was 1,730 μg/day (range of 27 to 10,500, SD = 2,350).  The average decrement in VASPI was 35 %; 30 out of 47 patients experienced a decrease greater than 30 % in visual analog scale of pain intensity (VASPI) while 13 of the 47 patients experienced a decrease greater than 50 % in VASPI.  There was no significant relationship between percent improvement in VASPI and morphine dosing; 8 of 47 combination patients experienced adverse events (AEs) attributable to intrathecal morphine but were capable to utilize the combination therapy for a least 1 year.  The author concluded that reduction in pain intensity with combined therapy was variable; intrathecal morphine can be a safe and effective adjunct pain therapy to patients utilizing intrathecal baclofen for spasticity.  The major drawbacks of this study were the lack of a control group and its relative small sample size (n = 47).

The above policy is based on the following references: