A quantitative evaluation of aerosol generation during cardiopulmonary resuscitation.

It is unclear if cardiopulmonary resuscitation is an aerosol-generating procedure and whether this poses a risk of airborne disease transmission to healthcare workers and bystanders. Use of airborne transmission precautions during cardiopulmonary resuscitation may confer rescuer protection but risks patient harm due to delays in commencing treatment. To quantify the risk of respiratory aerosol generation during cardiopulmonary resuscitation in humans, we conducted an aerosol monitoring study during out-of-hospital cardiac arrests. Exhaled aerosol was recorded using an optical particle sizer spectrometer connected to the breathing system. Aerosol produced during resuscitation was compared with that produced by control participants under general anaesthesia ventilated with an equivalent respiratory pattern to cardiopulmonary resuscitation. A porcine cardiac arrest model was used to determine the independent contributions of ventilatory breaths, chest compressions and external cardiac defibrillation to aerosol generation. Time-series analysis of participants with cardiac arrest (n = 18) demonstrated a repeating waveform of respiratory aerosol that mapped to specific components of resuscitation. Very high peak aerosol concentrations were generated during ventilation of participants with cardiac arrest with median (IQR [range]) 17,926 (5546-59,209 [1523-242,648]) particles.l-1 , which were 24-fold greater than in control participants under general anaesthesia (744 (309-2106 [23-9099]) particles.l-1 , p < 0.001, n = 16). A substantial rise in aerosol also occurred with cardiac defibrillation and chest compressions. In a complimentary porcine model of cardiac arrest, aerosol recordings showed a strikingly similar profile to the human data. Time-averaged aerosol concentrations during ventilation were approximately 270-fold higher than before cardiac arrest (19,410 (2307-41,017 [104-136,025]) vs. 72 (41-136 [23-268]) particles.l-1 , p = 0.008). The porcine model also confirmed that both defibrillation and chest compressions generate high concentrations of aerosol independent of, but synergistic with, ventilation. In conclusion, multiple components of cardiopulmonary resuscitation generate high concentrations of respiratory aerosol. We recommend that airborne transmission precautions are warranted in the setting of high-risk pathogens, until the airway is secured with an airway device and breathing system with a filter.

A quantitative evaluation of aerosol generation during awake tracheal intubation.

Aerosol-generating procedures are medical interventions considered high risk for transmission of airborne pathogens. Tracheal intubation of anaesthetised patients is not high risk for aerosol generation; however, patients often perform respiratory manoeuvres during awake tracheal intubation which may generate aerosol. To assess the risk, we undertook aerosol monitoring during a series of awake tracheal intubations and nasendoscopies in healthy participants. Sampling was undertaken within an ultraclean operating theatre. Procedures were performed and received by 12 anaesthetic trainees. The upper airway was topically anaesthetised with lidocaine and participants were not sedated. An optical particle sizer continuously sampled aerosol. Passage of the bronchoscope through the vocal cords generated similar peak median (IQR [range]) aerosol concentrations to coughing, 1020 (645-1245 [120-48,948]) vs. 1460 (390-2506 [40-12,280]) particles.l-1 respectively, p = 0.266. Coughs evoked when lidocaine was sprayed on the vocal cords generated 91,700 (41,907-166,774 [390-557,817]) particles.l-1 which was significantly greater than volitional coughs (p < 0.001). For 38 nasendoscopies in 12 participants, the aerosol concentrations were relatively low, 180 (120-525 [0-9552]) particles.l-1 , however, five nasendoscopies generated peak aerosol concentrations greater than a volitional cough. Awake tracheal intubation and nasendoscopy can generate high concentrations of respiratory aerosol. Specific risks are associated with lidocaine spray of the larynx, instrumentation of the vocal cords, procedural coughing and deep breaths. Given the proximity of practitioners to patient-generated aerosol, airborne infection control precautions are appropriate when undertaking awake upper airway endoscopy (including awake tracheal intubation, nasendoscopy and bronchoscopy) if respirable pathogens cannot be confidently excluded.

Anaesthetists' current practice and perceptions of aerosol-generating procedures: a mixed-methods study.

The evidence base surrounding the transmission risk of 'aerosol-generating procedures' has evolved primarily through quantification of aerosol concentrations during clinical practice. Consequently, infection prevention and control guidelines are undergoing continual reassessment. This mixed-methods study aimed to explore the perceptions of practicing anaesthetists regarding aerosol-generating procedures. An online survey was distributed to the Membership Engagement Group of the Royal College of Anaesthetists during November 2021. The survey included five clinical scenarios to identify the personal approach of respondents to precautions, their hospital's policies and the associated impact on healthcare provision. A purposive sample was selected for interviews to explore the reasoning behind their perceptions and behaviours in greater depth. A total of 333 survey responses were analysed quantitatively. Transcripts from 18 interviews were coded and analysed thematically. The sample was broadly representative of the UK anaesthetic workforce. Most respondents and their hospitals were aware of, supported and adhered to UK guidance. However, there were examples of substantial divergence from these guidelines at both individual and hospital level. For example, 40 (12%) requested respiratory protective equipment and 63 (20%) worked in hospitals that required it to be worn whilst performing tracheal intubation in SARS-CoV-2 negative patients. Additionally, 173 (52%) wore respiratory protective equipment whilst inserting supraglottic airway devices. Regarding the use of respiratory protective equipment and fallow times in the operating theatre: 305 (92%) perceived reduced efficiency; 376 (83%) perceived a negative impact on teamworking; 201 (64%) were worried about environmental impact; and 255 (77%) reported significant problems with communication. However, 269 (63%) felt the negative impacts of respiratory protection equipment were appropriately balanced against the risks of SARS-CoV-2 transmission. Attitudes were polarised about the prospect of moving away from using respiratory protective equipment. Participants' perceived risk from COVID-19 correlated with concern regarding stepdown (Spearman's test, R = 0.36, p < 0.001). Attitudes towards aerosol-generating procedures and the need for respiratory protective equipment are evolving and this information can be used to inform strategies to facilitate successful adoption of revised guidelines.

Quantitative evaluation of aerosol generation from upper airway suctioning assessed during tracheal intubation and extubation sequences in anaesthetized patients.

BACKGROUND: Open respiratory suctioning is defined as an aerosol generating procedure (AGP). Laryngopharyngeal suctioning, used to clear secretions during anaesthesia, is widely managed as an AGP. However, it is uncertain whether upper airway suctioning should be designated as an AGP due to the lack of both aerosol and epidemiological evidence. AIM: To assess the relative risk of aerosol generation by upper airway suctioning during tracheal intubation and extubation in anaesthetized patients. METHODS: This prospective environmental monitoring study was undertaken in an ultraclean operating theatre setting to assay aerosol concentrations during intubation and extubation sequences, including upper airway suctioning, for patients undergoing surgery (N=19). An optical particle sizer (particle size 0.3-10 µm) sampled aerosol 20 cm above the patient's mouth. Baseline recordings (background, tidal breathing and volitional coughs) were followed by intravenous induction of anaesthesia with neuromuscular blockade. Four periods of laryngopharyngeal suctioning were performed with a Yankauer sucker: pre-laryngoscopy, post-intubation, pre-extubation and post-extubation. FINDINGS: Aerosol was reliably detected {median 65 [interquartile range (IQR) 39-259] particles/L} above background [median 4.8 (IQR 1-7) particles/L, P<0.0001] when sampling in close proximity to the patient's mouth during tidal breathing. Upper airway suctioning was associated with a much lower average aerosol concentration than breathing [median 6.0 (IQR 0-12) particles/L, P=0.0007], and was indistinguishable from background (P>0.99). Peak aerosol concentrations recorded during suctioning [median 45 (IQR 30-75) particles/L] were much lower than during volitional coughs [median 1520 (IQR 600-4363) particles/L, P<0.0001] and tidal breathing [median 540 (IQR 300-1826) particles/L, P<0.0001]. CONCLUSION: Upper airway suctioning during airway management was not associated with a higher aerosol concentration compared with background, and was associated with a much lower aerosol concentration compared with breathing and coughing. Upper airway suctioning should not be designated as a high-risk AGP.

Quantitative evaluation of aerosol generation during manual facemask ventilation.

Manual facemask ventilation, a core component of elective and emergency airway management, is classified as an aerosol-generating procedure. This designation is based on one epidemiological study suggesting an association between facemask ventilation and transmission during the SARS-CoV-1 outbreak in 2003. There is no direct evidence to indicate whether facemask ventilation is a high-risk procedure for aerosol generation. We conducted aerosol monitoring during routine facemask ventilation and facemask ventilation with an intentionally generated leak in anaesthetised patients. Recordings were made in ultraclean operating theatres and compared against the aerosol generated by tidal breathing and cough manoeuvres. Respiratory aerosol from tidal breathing in 11 patients was reliably detected above the very low background particle concentrations with median [IQR (range)] particle counts of 191 (77-486 [4-1313]) and 2 (1-5 [0-13]) particles.l-1 , respectively, p = 0.002. The median (IQR [range]) aerosol concentration detected during facemask ventilation without a leak (3 (0-9 [0-43]) particles.l-1 ) and with an intentional leak (11 (7-26 [1-62]) particles.l-1 ) was 64-fold (p = 0.001) and 17-fold (p = 0.002) lower than that of tidal breathing, respectively. Median (IQR [range]) peak particle concentration during facemask ventilation both without a leak (60 (0-60 [0-120]) particles.l-1 ) and with a leak (120 (60-180 [60-480]) particles.l-1 ) were 20-fold (p = 0.002) and 10-fold (0.001) lower than a cough (1260 (800-3242 [100-3682]) particles.l-1 ), respectively. This study demonstrates that facemask ventilation, even when performed with an intentional leak, does not generate high levels of bioaerosol. On the basis of this evidence, we argue facemask ventilation should not be considered an aerosol-generating procedure.

Multisite silicon probes enable simultaneous recording of spontaneous and evoked activity in multiple isolated C-fibres in rat saphenous nerve.

BACKGROUND: Recordings of electrical activity in nerves have provided valuable insights into normal function and pathological behaviours of the nervous system. Current high-resolution techniques (e.g. teased fibre recordings) typically utilise electrodes with a single recording site, capturing the activity of a single isolated neuron per recording. NEW METHOD: We conducted proof-of-principle C-fibre recordings in the saphenous nerve of urethane-anaesthetised adult Wistar rats using 32-channel multisite silicon electrodes. Data was acquired using the OpenEphys recording system and clustered offline with Kilosort 2.5. RESULTS: In single recordings in 5 rats, 32 units with conduction velocities in the C-fibre range (< 1 m/s) were identified via constant latency responses and classified using activity dependent slowing. In two animals, 6 C-fibres (5 classified as nociceptors) were well isolated after clustering. Their activity could be tracked throughout the recording - including during periods of spontaneous activity. Axonal conduction velocities were calculated from spontaneous activity and/or low frequency electrical stimulation using only the differences in action potential latency as it propagated past multiple probe sites. COMPARISON WITH EXISTING METHODS: Single electrode approaches have a low data yield and generating group data for specific fibre types is challenging as it requires multiple experimental subjects and recording sessions. This is particularly true when the experimental targets are the small, unmyelinated C-fibres carrying nociceptive information. CONCLUSIONS: We demonstrate that multisite recordings can greatly increase experimental yields and enhance fibre identification. The approach is of particular utility when coupled with clustering analysis. Multisite probes and analysis approaches constitute a valuable new toolbox for researchers studying the peripheral nervous system.

A quantitative evaluation of aerosol generation during supraglottic airway insertion and removal.

Many guidelines consider supraglottic airway use to be an aerosol-generating procedure. This status requires increased levels of personal protective equipment, fallow time between cases and results in reduced operating theatre efficiency. Aerosol generation has never been quantitated during supraglottic airway use. To address this evidence gap, we conducted real-time aerosol monitoring (0.3-10-µm diameter) in ultraclean operating theatres during supraglottic airway insertion and removal. This showed very low background particle concentrations (median (IQR [range]) 1.6 (0-3.1 [0-4.0]) particles.l-1 ) against which the patient's tidal breathing produced a higher concentration of aerosol (4.0 (1.3-11.0 [0-44]) particles.l-1 , p = 0.048). The average aerosol concentration detected during supraglottic airway insertion (1.3 (1.0-4.2 [0-6.2]) particles.l-1 , n = 11), and removal (2.1 (0-17.5 [0-26.2]) particles.l-1 , n = 12) was no different to tidal breathing (p = 0.31 and p = 0.84, respectively). Comparison of supraglottic airway insertion and removal with a volitional cough (104 (66-169 [33-326]), n = 27), demonstrated that supraglottic airway insertion/removal sequences produced <4% of the aerosol compared with a single cough (p < 0.001). A transient aerosol increase was recorded during one complicated supraglottic airway insertion (which initially failed to provide a patent airway). Detailed analysis of this event showed an atypical particle size distribution and we subsequently identified multiple sources of non-respiratory aerosols that may be produced during airway management and can be considered as artefacts. These findings demonstrate supraglottic airway insertion/removal generates no more bio-aerosol than breathing and far less than a cough. This should inform the design of infection prevention strategies for anaesthetists and operating theatre staff caring for patients managed with supraglottic airways.

A quantitative evaluation of aerosol generation during tracheal intubation and extubation.

The potential aerosolised transmission of severe acute respiratory syndrome coronavirus-2 is of global concern. Airborne precaution personal protective equipment and preventative measures are universally mandated for medical procedures deemed to be aerosol generating. The implementation of these measures is having a huge impact on healthcare provision. There is currently a lack of quantitative evidence on the number and size of airborne particles produced during aerosol-generating procedures to inform risk assessments. To address this evidence gap, we conducted real-time, high-resolution environmental monitoring in ultraclean ventilation operating theatres during tracheal intubation and extubation sequences. Continuous sampling with an optical particle sizer allowed characterisation of aerosol generation within the zone between the patient and anaesthetist. Aerosol monitoring showed a very low background particle count (0.4 particles.l-1 ) allowing resolution of transient increases in airborne particles associated with airway management. As a positive reference control, we quantitated the aerosol produced in the same setting by a volitional cough (average concentration, 732 (418) particles.l-1 , n = 38). Tracheal intubation including facemask ventilation produced very low quantities of aerosolised particles (average concentration, 1.4 (1.4) particles.l-1 , n = 14, p < 0.0001 vs. cough). Tracheal extubation, particularly when the patient coughed, produced a detectable aerosol (21 (18) l-1 , n = 10) which was 15-fold greater than intubation (p = 0.0004) but 35-fold less than a volitional cough (p < 0.0001). The study does not support the designation of elective tracheal intubation as an aerosol-generating procedure. Extubation generates more detectable aerosol than intubation but falls below the current criterion for designation as a high-risk aerosol-generating procedure. These novel findings from real-time aerosol detection in a routine healthcare setting provide a quantitative methodology for risk assessment that can be extended to other airway management techniques and clinical settings. They also indicate the need for reappraisal of what constitutes an aerosol-generating procedure and the associated precautions for routine anaesthetic airway management.

Endogenous analgesic action of the pontospinal noradrenergic system spatially restricts and temporally delays the progression of neuropathic pain following tibial nerve injury.

Pontospinal noradrenergic neurons form part of an endogenous analgesic system that suppresses acute pain, but there is conflicting evidence about its role in neuropathic pain. We investigated the chronology of descending noradrenergic control during the development of a neuropathic pain phenotype in rats following tibial nerve transection (TNT). A lumbar intrathecal cannula was implanted at the time of nerve injury allowing administration of selective α-adrenoceptor (α-AR) antagonists to sequentially assay their effects upon the expression of allodynia and hyperalgesia. Following TNT animals progressively developed mechanical and cold allodynia (by day 10) and subsequently heat hypersensitivity (day 17). Blockade of α2-AR with intrathecal yohimbine (30 µg) revealed earlier ipsilateral sensitization of all modalities while prazosin (30 µg, α1-AR) was without effect. Established allodynia (by day 21) was partly reversed by the re-uptake inhibitor reboxetine (5 µg, i.t.) but yohimbine no longer had any sensitising effect. This loss of effect coincided with a reduction in the descending noradrenergic innervation of the ipsilateral lumbar dorsal horn. Yohimbine reversibly unmasked contralateral hindlimb allodynia and hyperalgesia of all modalities and increased dorsal horn c-fos expression to an innocuous brush stimulus. Contralateral thermal hyperalgesia was also reversibly uncovered by yohimbine administration in a contact heat ramp paradigm in anaesthetised TNT rats. Following TNT there is an engagement of inhibitory α2-AR-mediated noradrenergic tone which completely masks contralateral and transiently suppresses the development of ipsilateral sensitization. This endogenous analgesic system plays a key role in shaping the spatial and temporal expression of the neuropathic pain phenotype after nerve injury.

Is augmented central respiratory-sympathetic coupling involved in the generation of hypertension?

Respiratory modulation of autonomic neural activity, with consequent phasic alteration of cardiac and vascular function, has been observed in many species including humans and is considered an index of cardiovascular health. Whilst many factors contribute to this modulation, including for example baroreceptor reflex feedback, it is accepted that a significant component is derived from an interaction within the central nervous system. Functional links between the brainstem circuitry generating the respiratory rhythm and neurons responsible for generate sympathetic and parasympathetic activity to the cardiovascular system have long been hypothesized, although the detailed understanding of these interactions is incomplete. There are several proposed physiological functions for these interactions including the matching of ventilation to cardiac output and tissue blood flow. However, recent observations suggest that altered central respiratory coupling may play a role in the development of hypertension and in the maintenance of elevated levels of sympathetic vasomotor activity in disease. The focus of this review article is to discuss these observations and place them within the context of current understanding of the neural substrates that might be responsible for respiratory-sympathetic coupling.

The potency of different serotonergic agonists in counteracting opioid evoked cardiorespiratory disturbances.

Serotonin receptor (5-HTR) agonists that target 5-HT(4(a))R and 5-HT(1A)R can reverse mu-opioid receptor (mu-OR)-evoked respiratory depression. Here, we have tested whether such rescuing by serotonin agonists also applies to the cardiovascular system. In working heart-brainstem preparations in situ, we have recorded phrenic nerve activity, thoracic sympathetic chain activity (SCA), vascular resistance and heart rate (HR) and in conscious rats, diaphragmatic electromyogram, arterial blood pressure (BP) and HR via radio-telemetry. In addition, the distribution of 5-HT(4(a))R and 5-HT(1A)R in ponto-medullary cardiorespiratory networks was identified using histochemistry. Systemic administration of the mu-OR agonist fentanyl in situ decreased HR, vascular resistance, SCA and phrenic nerve activity. Subsequent application of the 5-HT(1A)R agonist 8-OH-DPAT further enhanced bradycardia, but partially compensated the decrease in vascular resistance, sympathetic activity and restored breathing. By contrast, the 5-HT(4(a))R agonist RS67333 further decreased vascular resistance, HR and sympathetic activity, but partially rescued breathing. In conscious rats, administration of remifentanyl caused severe respiratory depression, a decrease in mean BP accompanied by pronounced bradyarrhythmia. 8-OH-DPAT restored breathing and prevented the bradyarrhythmia; however, BP and HR remained below baseline. In contrast, RS67333 further suppressed cardiovascular functions in vivo and only partially recovered breathing in some cases. The better recovery of mu-OR cardiorespiratory disturbance by 5-HT(1A)R than 5-HT(4(a))R is supported by the finding that 5-HT(1A)R was more densely expressed in key brainstem nuclei for cardiorespiratory control compared with 5-HT(4(a))R. We conclude that during treatment of severe pain, 5-HT(1A)R agonists may provide a useful tool to counteract opioid-mediated cardiorespiratory disturbances.

Differential baroreflex control of sympathetic drive by angiotensin II in the nucleus tractus solitarii.

Microinjection of angiotensin II into the nucleus tractus solitarii attenuates the baroreceptor reflex-mediated bradycardia by inhibiting both vagal and cardiac sympathetic components. However, it is not known whether the baroreflex modulation of other sympathetic outputs (i.e., noncardiac) also are inhibited by exogenous angiotensin II (ANG II) in nucleus tractus solitarii (NTS). In this study, we determined whether there was a difference in the baroreflex sensitivity of sympathetic outflows at the thoracic and lumbar levels of the sympathetic chain following exogenous delivery of ANG II into the NTS. Experiments were performed in two types of in situ arterially perfused decerebrate rat preparations. Sympathetic nerve activity was recorded from the inferior cardiac nerve, the midthoracic sympathetic chain, or the lower thoracic-lumbar sympathetic chain. Increases in perfusion pressure produced a reflex bradycardia and sympathoinhibition. Microinjection of ANG II (500 fmol) into the NTS attenuated the reflex bradycardia (57% attenuation, P < 0.01) and sympathoinhibition of both the inferior cardiac nerve (26% attenuation, P < 0.05) and midthoracic sympathetic chain (37% attenuation, P < 0.05) but not the lower thoracic-lumbar chain (P = 0.56). We conclude that ANG II in the nucleus tractus solitarii selectively inhibits baroreflex responses in specific sympathetic outflows, possibly dependent on the target organ innervated.

A spinal vasopressinergic mechanism mediates hyperosmolality-induced sympathoexcitation.

An elevation in plasma osmolality elicits a complex neurohumoral response, including an activation of the sympathetic nervous system and an increase in arterial pressure. Using a combination of in vivo and in situ rat preparations, we sought to investigate whether hypothalamic vasopressinergic spinally projecting neurones are activated during increases in plasma osmolality to elicit sympathoexcitation. Hypertonic saline (HS, i.v. bolus), which produced a physiological increase in plasma osmolality to 299 +/- 1 mosmol (kg water)(-1), elicited an immediate increase in mean arterial pressure (MAP) (from 101 +/- 1 to 121 +/- 3 mmHg) in vivo. Pre-treatment with prazosin reversed the HS-induced pressor response to a hypotensive response (from 121 +/- 3 to 68 +/- 2 mmHg), indicating significant activation of the sympathetic nervous system. In an in situ arterially perfused decorticate rat preparation, hyperosmotic perfusate consisted of either 135 mm NaCl, or a non-NaCl osmolyte, mannitol (0.5%); both increased lumbar sympathetic nerve activity (LSNA) by 32 +/- 5% (NaCl) and 21 +/- 1% (mannitol), which was attenuated after precollicular transection (7 +/- 3% and 1 +/- 1%, respectively). Remaining experiments used the NaCl hyperosmotic stimulus. In separate preparations the hyperosmotic-induced sympathoexcitation (21 +/- 2%) was also significantly attenuated after transection of the circumventricular organs (2 +/- 1%). Either isoguvacine (a GABA(A) receptor agonist) or kynurenic acid (a non-selective ionotropic glutamate receptor antagonist) microinjected bilaterally into the paraventricular nucleus (PVN) attenuated the increase in LSNA induced by the hyperosmotic stimulus (control: 25 +/- 2%; after isoguvacine: 7 +/- 2%; after kynurenic: 8 +/- 3%). Intrathecal injection of a V(1a) receptor antagonist also reduced the increase in LSNA elicited by the hyperosmotic stimulus (control: 29 +/- 6%; after blocker: 4 +/- 1%). These results suggest that a physiological hyperosmotic stimulus produces sympathetically mediated hypertension in conscious rats. These data are substantiated by the in situ decorticate preparation in which sympathoexcitation was also evoked by comparable hyperosmotic stimulation. Our findings demonstrate the importance of vasopressin acting on spinal V(1a) receptors for mediating sympathoexcitatory response to acute salt loading.

The yin and yang of cardiac autonomic control: vago-sympathetic interactions revisited.

We review the pattern of activity in the parasympathetic and sympathetic nerves innervating the heart. Unlike the conventional textbook picture of reciprocal control of cardiac vagal and sympathetic nervous activity, as seen during a baroreceptor reflex, many other reflexes involve simultaneous co-activation of both autonomic limbs. Indeed, even at 'rest', the heart receives tonic drives from both sympathetic and parasympathetic cardiac nerves. Autonomic co-activation occurs during peripheral chemoreceptor, diving, oculocardiac, somatic nociceptor reflex responses as well as being evoked from structures within the brain. It is suggested that simultaneous co-activation may lead to a more efficient cardiac function giving greater cardiac output than activation of the sympathetic limb alone; this permits both a longer time for ventricular filling and a stronger contraction of the myocardium. This may be important when pumping blood into a constricted vascular tree such as is the case during the diving response. We discuss that in some instances, high drive to the heart from both autonomic limbs may also be arrhythmogenic.

Double-blind, placebo-controlled analgesic study of ibuprofen or rofecoxib in combination with paracetamol for tonsillectomy in children.

BACKGROUND: The analgesics used for paediatric tonsillectomy may be associated with side-effects such as sedation, respiratory depression and vomiting (opioids) or increased bleeding [non-steroidal anti-inflammatory drugs (NSAIDs)]. In our institution, we employ a combination of paracetamol, NSAID and opioid, although there is no published evidence of analgesic benefit from adding NSAIDs to paracetamol in children. METHODS: This randomized, double-blinded clinical study examined the analgesic effectiveness of combining paracetamol (20 mg kg(-1)) with rofecoxib (0.625 mg kg(-1)), ibuprofen (5 mg kg(-1)) or placebo as premedication for (adeno)tonsillectomy (n=98) in children aged 3-15 yr. Intravenous fentanyl 1-2 microg kg(-1) was given intraoperatively. Regular oral paracetamol (15 mg kg(-1), 4 hourly) was given after operation and could be supplemented on request from the child with oral ibuprofen 5 mg kg(-1) or oral codeine 1 mg kg(-1). The primary outcome variable was need for early supplementary analgesia (within 2 h after surgery). RESULTS: The addition of ibuprofen to paracetamol reduced the need for early analgesia from 72% to 38% of children (difference 34%; 95% confidence interval 4-64%). The addition of rofecoxib to paracetamol did not significantly alter the need for early analgesia (68 vs 72%). Pain scores were higher in those children who required early analgesia. There were no differences between the groups in operative blood loss or complications, total 24-h analgesic consumption, pain scores at 4 and 8 h, vomiting or antiemetic use. CONCLUSION: This study provides evidence to support the combination of ibuprofen (but not rofecoxib) with paracetamol for perioperative analgesia in children.