The intrathecal delivery of drug (ITD) has clear theoretical advantages, as it bypasses the blood-brain barrier and allows for an increase in medication potency of an order of magnitude. Clinically ITD has been shown to be more effective than oral administration in the treatment of intractable spasticity.1 There is however no well controlled study to demonstrate the superior pain relief of ITD over systemic administration in chronic non-cancer pain patients.2 One large multicentre randomised controlled trial in end-stage cancer patients, comparing ITD with the ‘best current medical treatment’ has shown a significant decrease in side effects in patients receiving ITD. There was however no significant difference in the intensity of pain.3 Despite appealing features, the clinical benefit of ITD is hampered by a number of biological, pharmacological and organisational problems.4 Even the most enthusiastic supporters will agree that this treatment is far more complicated and resource-demanding than the oral route. As with any surgery, there is a modest risk of infection, but meningitis and encephalitis are major, potentially life-threatening complications. Other surgical problems include pocket, epidural or spinal haematomas.5 The choice of drugs to be used with intrathecal pumps is limited and only baclofen and ziconotide have been unanimously recognised. the FDA has accepted morphine but not the EMA, and in Europe, the use of morphine depends on the decision of each countries. To complicate matters, the overwhelming clinical experience suggests that IT morphine is often insufficient to achieve pain relief in the most difficult pain patients, particularly when neuropathic and mechanical components are predominant. Yet the data shows the majority of clinicians do use a large number of other drugs,6 which means that intrathecal administration of drugs is mostly performed off label7. Though compounding is endorsed by an international consensus, the theoretical risk of litigation remains. Finally, the provision of ITD cannot be done by one physician but requires a team of doctors and specifically trained nurses. Patients have to be seen in the clinic at regular intervals to refill the pump, which necessitates an infrastructure to ensure that patients are followed and called in time. Furthermore, an on-call system has to be put in place as technical problem (such as pump stalls) can occur and have the potential of precipitating severe withdrawal symptoms that may not necessarily be recognised by a medical team unfamiliar with ITD. In addition, the pump refill procedure requires trained personnel as it carries a risk of overdose either because of an error in the selection of the Injection site (site port instead of pump reservoir), or more frequently pocket site injection.8 9 In conclusion the efficacy of ITD for the management of severe pain is essentially supported by personal experience and limited uncontrolled case series and is therefore still debatable. the fact remains that no controlled trial has documented its superiority over oral administration, whether in cancer or non-cancer pain. Admittedly, the administration of drugs such as local anaesthetics and clonidine are not possible through a systemic route and may be necessary in severe mechanical and/or neuropathic pain. But theses are relatively uncommon situations that do not justify the widespread use of this technique. References Coffey JR, et al. Intrathecal baclofen for intractable spasticity of spinal origin: results of a long-term multicenter study. J Neurosurg 1993;78:226–232. Turner JA, Sears JM, Loeser JD. Programmable intrathecal opioid delivery systems for chronic noncancer pain: a systematic review of effectiveness and complications. Clin J Pain 2007;23:180–195. Smith TJ, et al. Randomized clinical trial of an implantable drug delivery system compared with comprehensive medical management for refractory cancer pain: Impact on pain, drug-related toxicity, and survival. J Clin Oncol 2002;20:4040–4049. doi:10.1200/Jco.2002.02.118 Prager J, et al. Best practices for intrathecal drug delivery for pain. Neuromodulation 2014;17:354–372; discussion 372. doi:10.1111/ner.12146 Deer TR, et al. The Polyanalgesic Consensus Conference (PACC): Recommendations for Intrathecal Drug Delivery: Guidance for Improving Safety and Mitigating Risks. Neuromodulation 2017;20:155–176. doi:10.1111/ner.12579 Hassenbusch SJ, Portenoy RK. Current practices in intraspinal therapy-a survey of clinical trends and decision making. J.Pain Symptom.Manage. 2000.Aug;20(2):S4–11. 2000;20:S4–11. Deer TR, et al. The Polyanalgesic Consensus Conference (PACC): Recommendations on Intrathecal Drug Infusion Systems Best Practices and Guidelines. Neuromodulation 2017;20:96–132. doi:10.1111/ner.12538 Maino P, Perez R, Koetsier E. Intrathecal pump refills, pocket fills, and symptoms of drug overdose: a prospective, observational study comparing the injected drug volume vs. the drug volume effectively measured inside the pump. Neuromodulation 2017;20:733–739. doi:10.1111/ner.12597 Wankhede V, Buchser E. a case of pocket fill causing baclofen withdrawal. Neuromodulation 2018. doi:10.1111/ner.12887