This study shows that INKM administered via MAD seems to be effective and safe when gastric washings in anxious and uncooperative children need to be performed to diagnose TB. All patients achieved an adequate level of sedation throughout each procedure, with a statistically significant difference compared to the placebo-group in the degree of MOPS reduction. Both the medical staff and the patient’s parents reported a significant degree of satisfaction with the INKM protocol used to perform gastric aspirates respect to the placebo group. There were no serious side effects, such as oxygen desaturation, bradycardia, hypotension or apnoea. The only side effect registered was post-sedation agitation. As described in literature, the use of Ketamine is known to result in an agitated, confused, and combative child. There has been no documented effective strategy in the literature to manage this problem but these adverse reactions are usually transient and completely recover spontaneously without any adverse consequences [
21]. On the other hand, several studies revealed that intranasal Midazolam administered by MAD or by drops resulted in nasal burning and a bitter taste in up to 66% of patients, making the experience unpleasant and the procedure more difficult to complete [
22,
23]. The administration of intranasal Lidocaine prior to the use of intranasal Midazolam has been reported to be beneficial in reducing the burning sensation from the intranasal Midazolam [
11]. In our experience we decided to not use intranasal Lidocaine before the Midazolam administration due to the local anesthetic properties of the Ketamine able to avoid any nasal discomfort related to intranasal Midazolam. Recent studies also reported the use of much higher doses of intranasal Ketamine (up to 9 mg/kg) to perform different painful procedures in pediatric age [
24,
25]. In our experience we decided to utilize a relatively low dose of Ketamine to evaluate both its efficacy and also to provide a safety dose of the drug able to avoid the more frequent side effects related to its use, such as nausea, vomiting, respiratory depression, and laryngospasm. Recently, it has been reported that high doses of Ketamine in different procedures involving major airways, such as gastroscopy and bronchoscopy, are responsible of the appearance of the laryngospasm in about 10% of cases needed O
2 ventilation by face-mask and continuous positive airway pressure to be resolved [
11,
26,
27]. Our results showed that, despite the low doses of intranasal Ketamine, we were able to complete all the procedures without major side effects, such as laryngospam, in treated children making this procedure very easy. In fact, the only side effect observed in the sedation group was the post-sedation agitation in 14.3% of cases. Obviously, a slightly higher dose of Ketamine could have induced a higher sedation allowing to complete the procedure by a single physician in every cases, particularly in younger children, since other authors found that younger children required higher dosage in milligrams per kilogram of Ketamine for adequate sedation [
28]. This may be explained by the faster metabolism and renal clearance leading to a shorter half-life of Ketamine in children compared to adults [
29,
30]. Nevertheless, Hosseini Jahromi SA et al. recently found that by increasing the dose of intranasal Ketamine (0.5 mg/kg vs 3 mg/kg), the children developed less sedation than those who received a lower dose of Ketamine [
25], suggesting that a dosage as low as 0.5 mg/kg intranasal Ketamine might be suitable for achieving an adequate therapeutic action (sedation and anxiolysis) combined with lesser side effects because there is a dose–response correlation between Ketamine and its side effects [
8,
14,
28]. Similarly, Yeaman F et al. recently demonstrated that sub-dissociative doses of intranasal Ketamine provided adequate analgesia by 30 min for most paediatric patients aged 3-13 years admitted to the emergency department with moderate to severe pain from isolated limb injury [
31]. The PICHFORK (Pain In CHildren Fentanyl OR Ketamine) trial [
32] aimed to compare the efficacy of intranasal Ketamine and Fentanyl in the relief of moderate to severe pain in children with limb injuries will use low doses of intranasal Ketamine (1 mg/kg). All these findings support our choice of using moderate doses of Ketamine (2 mg/kg) in order to achieve a proper sedative/analgesic effect while reducing the possibility of severe adverse events, making our protocol applicable also in pediatric wards. Moreover, in our study, we did not applied local anesthetics on nasogastric tube, for two reasons. Firstly, in current literature it is not recommended for the specific use of collect samples for the detection of
Mycobacterium tuberculosis (MTB) in children [
18]; secondly, local anesthetics have a well established antimicrobial activity versus bacteria and fungi [
33‐
35]. Although this aspect has not yet been studied versus mycobacteria, the even low risk of potentially reducing the possibility of culture MTB induced us to not apply local anesthetics on nasogastric tubes. Moreover, even though local anesthetics may reduce the pain related to NGTI, it would not change the experience of the child undergoing the procedure. This is particularly important for those cases, such as suspected TB patients, who need to be submitted to frequent NGTIs to perform the diagnosis and for treatment monitoring. The advantage of a procedural sedation over the local anesthetic, in such situations, is also to improve the all experience for both doctors, physicians and parents particularly when a single child need to undergo this procedure more times. Our results confirm all those findings making our INKM-based protocol a simple, safe and affordable way to perform this procedure in uncooperative children. The use of this protocol would allow even a single nurse or physician (after a brief training program) to perform the procedure, ensuring pain relief and patients and parents’ comfort. Our results could be particularly useful in low-income high TB-burden countries, where a big number of gastric washings would need to be performed daily and where there are limited physician and only a few with specific training in procedural sedation [
36]. Moreover, our protocol had low drug-related costs and allowed to reduce the number of health care workers necessary to perform the procedure. Nevertheless, these savings should be counterbalanced with costs related to monitoring (both health care workers and equipment). In fact, we found a long median recovery time (79.7 min). The recovery time has wide individual variations and depends on the doses of medications and clearance in the body. However, the relatively long recovery time is not a major problem in ward settings and day hospital regimens.
Limitations of the study
The first one is the low number of children evaluated. This makes the data difficult to interpret and did not allow us to find any variables (gender, weight, ethnicity, etc) that could be associated with a higher degree of sedation. Moreover, the low number of children did not allow us to strongly affirm the safety of our protocol, although the preliminary results of our study are promising. If our results would be confirmed by a multicentre, prospective, controlled, randomized trial our protocol could be easily introduced in general practice not only for gastric washings in children with suspected TB but also for different kind of procedural sedation in uncooperative children.
Secondly, the subjectivity of pain measurement by both parents and physicians is a clear limitation which is difficult to overcome. Several sedation scoring scales have been described for children but in current literature there are no comparison studies of all scales; nevertheless, the validity of MOPS has been clearly described [
19,
20]. In this regard, other pain scores have been described in literature. Among those, the FLACC scale have been widely used for post-operative and non post-operative pain assessment, and could have been used also in our study [
37]. Nevertheless, we decided to not use FLACC score because the Authors clearly showed that this score can be high during non-painful procedures and during restraint phase of painful procedures, indicating that FLACC scale measures a composite of pain and distress in young children, in this way altering the application of this score in our study [
38]. Finally, we analyzed drug-related costs, which are affordable even in low income countries and allow to perform procedures by a lower number of health care workers, but clearly need to be counterbalanced by monitoring and equipment costs, making difficult to establish the cheapest way to perform these procedures. On the other hand, parents appreciation appeared clear. Therefore, the use of the protocol should be evaluated depending on local resources.