INTRODUCTION
The phrenic nerve originates from the cervical nerve roots C3-C5 and innervates the diaphragm, the primary muscle responsible for inspiration [1]. Dysfunction or pathology affecting the diaphragm can lead to debilitating symptoms, ranging from dyspnea - shortness of breath - and respiratory to non-respiratory manifestations such as hiccups and shoulder pain, respectively. While not yet widely employed in clinical practice, ultrasound-guided phrenic nerve blocks (UPNBs) may have therapeutic benefits in various clinical scenarios. In 2002, Calvo et al. [2] were the first to publish the use of UPNB in clinical practice, demonstrating its efficacy in ameliorating intractable hiccups in cancer patients. Since then, UPNB has shown efficacy in enhancing lung re-expansion after pneumothorax [3]; reducing shoulder pain after hepatectomy [4], cholecystectomy [5], thoracotomy [6], or diaphragmatic surgery [7,8]; and assistance for lung biopsy [9,10].
While landmark-based phrenic nerve blockade has also been utilized, it is associated with higher rates of documented complications such as iatrogenic pneumothorax as well as potential vascular injury given the adjacent position of the carotid artery and jugular veins [11]. The ultrasound-guided technique offers a precise and real-time approach to localize and target the phrenic nerve, facilitating accurate administration of therapeutic agents close to its course. Given that ultrasound enhances the clinical safety of regional anesthesia, we chose to focus on UPNB in this review instead of landmark-based techniques. The UPNB can be performed by identifying the interscalene brachial plexus on ultrasound (Fig. 1). Instead of directing the needle through the middle scalene to the sheath containing the C5 through C7 nerve roots as in the interscalene brachial plexus block (ISB), the physician directs the needle in a much shallower trajectory towards the phrenic nerve, which lies within the fascial plane between the sternocleidomastoid and anterior scalene muscles, medial and superficial to the brachial plexus nerve roots. Successful blockade can be confirmed with ultrasound demonstrating reduced or absent excursion of the right hemidiaphragm, a reduction in pulmonary function testing (PFT) parameters, or simply by ultrasound visualization of the phrenic nerve within the fascial plane bathed in anesthetic [12].
Given the central role of the phrenic nerve in providing motor and sensory innervation to the ipsilateral hemidiaphragm and thus its importance in respiratory function, it would logically follow that UPNBs may lead to respiratory decompensation in spontaneously breathing patients.
However, appropriately selected patients may tolerate transient hemidiaphragmatic paralysis, likely due to compensatory effects of the contralateral diaphragm as well as “accessory” muscles of respiration (including the intercostals, sternocleidomastoid, and scalene muscles) [13]. Despite UPNB techniques growing increasingly popular in the last 15 years in regional anesthesia, they have been commonly observed as an unintentional outcome of an ISB, instead of a thoughtful procedure [14]. Bergmann and colleagues note that phrenic nerve block is in most cases unintended consequences of ISB due to reduction in forced expiratory volume and inspiratory strength [14]. In this systematic review, therefore, we evaluate the respiratory effects of UPNBs in spontaneously breathing patients and discuss their effectiveness in managing respiratory conditions and metrics.
METHODS
Study design
We performed this systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [15]. Two authors (MS and CR) searched the International Prospective Register of Systematic Reviews (PROSPERO) to confirm that no previous systematic review had been performed on the same subject. We registered the protocol on PROSPERO (ID: CRD42024501299).
Literature search
We developed a search protocol and independently conducted a systematic search across MEDLINE via PubMed, Embase, Web of Science, and Google Scholar to identify relevant studies published up to May 2025. The search strategy utilized Medical Subject Headings (MeSH) terms and a combination of keywords, including “respiratory effects,” “ultrasound-guided phrenic nerve block,” and “ultrasound phrenic nerve block.” A detailed search strategy for each database is shown in the supplementary file (Supplementary Table 1). The authors also manually retrieved all publications within citations and bibliographies of studies identified in this search.
Eligibility criteria
1. Inclusion criteria
Studies were eligible for inclusion if they met the following conditions:
(1) Studies that described phrenic nerve blockade performed under real-time ultrasound guidance, alone or in conjunction with nerve stimulation or other nerve blocks.
(2) Studies that specifically included UPNB performed either in awake patients or intraoperatively with longer-lasting anesthetic shortly before extubation and expected to affect patients while awake and breathing spontaneously.
2. Exclusion criteria
Studies were excluded due to the following reasons:
(1) Studies that had blockade performed either by landmark methods or nerve stimulator alone without ultrasound guidance; phrenic nerve ablation, ligation, or clipping (reversible or irreversible).
(2) Studies describing only intra-operative phrenic nerve blockade which did not affect awake patients.
(3) Studies which did not describe the amount and concentration of anesthetic used or studies of cryolysis.
(4) Secondary studies, including reviews and meta-analyses.
(5) Non-full text studies such as abstracts.
(6) Studies published in languages other than English.
(7) Non-human or cadaveric studies.
All citations were imported into Google Sheets (Google LLC, USA). After completing their search, the two authors met to discuss finalized inclusion. The third author (TZ) was available in the case that a finalized consensus could not be reached.
Study selection
The search results were imported into the Google Sheet for criteria assessment. All of the titles and abstracts that were retrieved were first reviewed by one reviewer (MS) to eliminate duplicates, and then they were included in the review based on the screening criteria. To ascertain if a study qualified for inclusion in the whole analysis, the full texts of prospective articles were acquired and assessed by two reviewers (MS and CR). Publications pertaining to the same study were checked for overlapping data, and the most pertinent report (based on study results) was chosen.
Data extraction and outcomes
Two authors (TZ and MS) were independently tasked with extracting essential data from the studies that met the inclusion criteria for this review. The vital information extracted was: study ID (author name[s] and publication year), study design, patient characteristics (sample size of patients who received UPNB, age, body mass index [BMI], and sex), and primary (minimized dyspnea [shortness of breath] or desaturation [decrease in oxygen saturation] after UPNB) and secondary outcomes (changes in arterial blood gas [ABG] samples, PFT), reduction in intractable/persistent hiccups, pneumothorax). PFTs involved assessing peak expiratory flow (PEF), forced expiratory volume in one second (FEV1), and forced vital capacity (FVC). Conversely, ABG sample analysis involved recording changes in blood pH (acid-base balance), PaO2 (partial pressure of oxygen), PaCO2 (partial pressure of carbon dioxide), and SaO2 (blood oxygen saturation). Anesthetic type, amount, concentration, and adjuvants were recorded. For studies that did not include exact ages or only the age by the decade, the age was rounded down to the lowest number (i.e., “a male in his 60s” was reported as one male aged 60). Any discrepancies between the two authors during the data extraction process were resolved through consensus and discussion.
Assessment of methodological quality of individual studies
Due to the variability of study designs for the included studies, we utilized different tools for quality assessment. The revised Cochrane risk of bias (RoB 2) tool was used to assess the risk of bias among included randomized controlled trials (RCTs). One reviewer (TZ) was tasked with assessing the risk of bias among included RCTs. Conversely, we used the Newcastle-Ottawa Scale for observational studies. Additionally, the Joanna Briggs Institute (JBI) critical appraisal checklists for case series and case reports were used to assess the methodological quality of case series and case reports, respectively [16]. Two reviewers (TZ and MS) assessed the methodological quality of the observational studies, case reports, and case series. Any inconsistencies regarding the methodological quality of each study were resolved through discussion. If solution was not met, the third reviewer (CR) was consulted.
RESULTS
Search outcomes
A total of 678 studies that we deemed potentially eligible were identified in the literature search. Before reviewing, we deduced 196 duplicates and thus, were excluded. The remaining 482 articles were examined based on title and abstract screening to determine irrelevant studies. Consequently, we excluded 243 studies as they were irrelevant to our study objective. We sought the remaining 239 records for retrieval. We did not retrieve 12 studies (reason: they were ongoing studies yet to provide outcomes on patients) and only 227 articles were available for full-text eligibility assessment. Only 26 articles met the inclusion criteria and were selected for this review. The remaining 201 articles were excluded due to the following reasons: 17 assessed non-human subjects such as pigs, 48 had blockade performed either by landmark methods or nerve stimulator alone without ultrasound guidance, 46 did not specify anesthetic concentration or amount used among patients, 40 targeted blockades other than phrenic nerve blockade (such as brachial plexus block), 31 were book chapters, and 19 described only intra-operative phrenic nerve blockade which did not affect awake patients. Fig. 2 summarizes the identification of studies via database searches.
Patient and study characteristics
Table 1 below details the characteristics of the included studies. Among the included studies, one was retrospective [10], three RCTs [4-6], eighteen case reports [3,7,8,17-30,31], and four case series [2,9,12,32]. Across all combined studies, 188 patients received a UPNB. Sixty-eight (36.2%) were females and 115 (61.2%) were males. Five patients did not have sex documented. The mean age for patients across the selected studies was 64.3 years old. The median BMI was 24.2 (range 22.0 to 28.9). BMI data were missing in 95 patients (50.5%). Study dates ranged from 2002 to 2024.
Outcomes
1. Primary outcomes
Eight studies highlighted UPNB’s effect in attaining/maintaining standard oxygen saturation rate (no desaturation) among patients after phrenic nerve block [4,5,18,19,21,31,32]. Bak et al. [4] and Patella et al. [12] postulate that phrenic nerve block had no cases of desaturation. For the patient assessed by Zhang et al. [21], he attained a 99% oxygen saturation after the nerve block. Similarly, Gong et al. [32] state that after phrenic nerve block, patients attained oxygen saturation of between 96-100%, which are excellent figures. On the other hand, one of the cases assessed by Thanaboriboon et al. [31] maintained her oxygen saturation after undergoing right-sided UPNB.
Only three studies reported outcomes on dyspnea [9,25,32]. Carrero et al. [9] highlight that after the nerve block, the two patients recovered with no cases of dyspnea. Le-Wendling et al. [25] stress that injecting phrenic nerve blockade did not worsen subjective dyspnea among patients with shortness of breath. Besides, Gong et al. [32] state that after one month of applying the nerve block, patients reported no severe complications, including dyspnea.
2. Secondary outcomes
Three studies reported reduced diaphragmatic pain after application of UPNB among patients [7,8,31]. Biswas et al. [7] highlight that the nerve block relieved the patient’s sensation of fullness and pulling in the epigastric region, contributing to complete diaphragmatic pain reduction. Similarly, Dontukurthy et al. [8] highlight a similar outcome: the patient exhibited diaphragmatic pain relief from 8/10 to 3/10 on a visual analog scale after a nerve block. Thanaboriboon et al. [31] note that all two cases attained significant diaphragmatic pain reduction. The outcome had effect on reduction of shoulder and upper abdominal pain.
Assessing PFTs (FVC, FEV1, and PEF) was available in 79 patients (42.5%) among four studies [4-6,22]. Bak et al. [4] note no significant decrease in PFTs when comparing pre-operative to post-operative due to block effect. Specifically, FVC, PEF, and FEV1 remained statistically the same after block, indicating no nerve block side effects on PFTs. Similarly, the authors note that the ABG sample, particularly PaCO2, remains statistically the same after UPNB with no cases of hypercapnia [4]. Blichfeldt-Eckhardt et al. [6] note that there were no differences in PFTs (mean FEV1 and FVC) after the nerve block. Blichfeldt-Eckhardt et al. [6] also highlight that ABG samples remained statistically the same after nerve block, with no PaO2, PaCO2, and arterial pH reduction. Yi et al. [5] posit that after the UPNB, patients exhibited decreased PFT values but normalized after 24 h. The authors note no significant differences between FEV and PEF values pre- and post-block. Nevertheless, there was a significant difference in FVC [5]. Renes et al. [22] postulate that spirometry (PFT) values FEV1, FVC, and PEF moderately reduced. Specifically, FEV1 (2.25 L) and FVC (2.36 L) reduced to FEV1 (1.98 L) and FVC (2.05 L).
Besides, the review highlights that after applying phrenic nerve block, patients’ intractable or persistent hiccups reduced significantly or ceased altogether, with 15 studies reporting this outcome [2,17-24,26-32]. For instance, Süren et al. [17] explain that after UPNB, the patients experienced a 70% reduction in hiccup severity and frequency.
Moreover, three studies reported that nerve block influences a significant reduction of pneumothorax among patients [3,9,10]. Mishra et al. [3] note that the patient experienced a significant decrease in pneumothorax in 24 h compared to immediate post-operative nerve block. Five days later, the patient attained complete pneumothorax resolution. Carrero et al. [9] and Czaplicki et al. [10] explain that the application of UPNB ceased the progression of pneumothorax among patients.
Methodological quality of the included studies
The 18 case reports’ quality was generally good, with all scoring “yes” for most domains. The patient’s domain history was poorly illustrated in all the reports. Patient’s related history is not described in detail with lack of structured timeline presentation. The detailed quality assessment for each primary record can be found in the supplementary materials (Supplementary Table 2). For the included RCTs, only two studies [5,6] had an overall low risk of bias, while one [4] had overall “Some concerns” (Fig. 3). The study by Bak et al. [4] had some concerns under ‘Missing outcome data’ domain. The authors did not provide explicit data on missing outcome data, specifically, follow-up attrition regarding patients’ follow-up on assessment of pain scores.
For one included retrospective study, its methodological quality assessed by the Newcastle-Ottawa Scale was fair (Supplementary Table 3). Supplementary Table 4 displays the outcomes of the methodological quality assessment of the case series included in this review. Most case series had overall fair quality. The variability of the study designs (most being case reports and case series) hindered conducting a meta-analysis. Similarly, the included RCTs had a variance in reporting the outcomes, preventing quantitative analysis.
DISCUSSION
We performed this systematic review without quantitative analysis due to the limitation of case reports and case series being the majority of included studies. The findings show that UPNB does not induce or worsen dyspnea among patients. Patients who used the technique did not experience desaturation but maintained or improved oxygen saturation, with some studies reporting 99% and 96-100% oxygen saturation [21,32]. Besides, phrenic nerve block injection is vital in reducing diaphragmatic pain. Patients who exhibited intractable or persistent hiccups had their frequency and severity reduced or vanished after the nerve block injection. Other outcomes that UPNB affected were ABG samples (PaO2, PaCO2, and arterial pH) and spirometry/PFT values (FEV1, FVC, and PEF). Most studies report that UPNB did not adversely affect the ABG samples and spirometry values, suggesting its lack of adverse impacts among patients.
The abdominal musculature, usually responsible for expiration, can adapt to become drivers of inspiration, when necessary, such as in diaphragmatic paresis [33]. Furthermore, none of these “accessory” muscles of inspiration are innervated by the phrenic nerve. Therefore, in hemidiaphragmatic paralysis, the intercostals, scalenes, sternomastoids, abdominal musculature, and even the contralateral hemidiaphragm assume the task of inspiration and ventilation [33]. In fact, even in bilateral diaphragmatic paralysis, accessory muscles may allow for normal respiration and oxygenation: Süren et al. [17] performed bilateral UPNBs simultaneously in a patient with intractable hiccups who subsequently did not desaturate. While our data are limited by the total patient number (186), no study reported cases of dyspnea or shortness of breath (desaturation) after UPNB.
On the contrary, despite the compensatory role of accessory muscles of inspiration after hemidiaphragmatic paresis, it has been well established that particular patients who experience inadvertent phrenic nerve blockade, such as from an interscalene block, are at high risk for decompensation [34]. Sripiya et al. [35] observed that the phrenic nerve courses very close to the C5 ventral ramus, usually within 2 mm on ultrasound measurements. Thus, it is likely that diffusion of local anesthetic across the anterior scalene muscle itself or within the fascial plane between the sternocleidomastoid and the anterior scalene muscles during an interscalene block results in an unintended phrenic nerve blockade. Patients who cannot withstand a significant decrease in inspiratory function, such as those with chronic obstructive pulmonary disease, asthma, obstructive sleep apnea, morbid obesity, American Society of Anesthesiologists class III or higher, and congestive heart failure, have been shown to have significantly higher rates of respiratory complications secondary to inadvertent phrenic nerve blockade [36-38]. Though exclusion criteria were not documented in all included studies in this review, patients with the aforementioned comorbidities were often excluded when criteria were available. While BMI data were also missing from several patients, only two 186 had a documented BMI over 25. Therefore, this review does not indicate a role for UPNB within previously outlined high-risk patients.
There are several reasons to target only the phrenic nerve, not the entire interscalene brachial plexus. Firstly, the interscalene block involves a target closer to the subclavian and vertebral arteries. Thus, intravascular injection and hematoma formation are theoretically more likely to occur with an interscalene block than with a UPNB. Furthermore, the dorsal scapular and long thoracic nerves may lie within the body of the middle scalene muscle and thus may course within the needle trajectory of an interscalene block. Injury to these nerves may lead to winged scapula [34].
Additionally, the anatomy of the interscalene brachial plexus may predispose to other neurologic injuries: the injection of anesthetic at the division of either the C5 or C6 nerve roots can lead to injury of the roots themselves, and the spread of anesthetic to the neuraxial space [39]. Physicians often mistake the “stoplight” target of the interscalene block for the C5, C6, and C7 nerve roots, when in fact, it may represent C5 and the division of C6 with C7 much deeper and nearer to the vertebral artery [40]. Holbrook and Parker [41] published a systematic review of interscalene blocks performed before shoulder surgery. They demonstrated that greater than 3% of patients who received a single-shot interscalene block suffered from neurologic injury postoperatively, which persisted after one year in 0.5% of patients. Other potential neurologic complications of the interscalene block include hoarseness secondary to paralysis of the ipsilateral vocal cords from injury to the recurrent laryngeal nerve and Horner’s syndrome (ptosis, miosis, and anhidrosis) from injury to the stellate ganglion [42]. Therefore, since the interscalene block harbors its risks, some severe or permanent targeted phrenic nerve blockade may reduce the risk of these complications. The phrenic nerve is a shallow target within a fascial plane between the sternocleidomastoid and anterior scalene muscles. Since it is a plane block, the UPNB can potentially allow the physician to decrease the risk of nerve injury further by keeping the needle more distant from the phrenic nerve.
LIMITATIONS
The included studies mainly included case studies and case series, limiting the quantitative outcomes analysis. The publications analyzed in this review did not contain consistent information regarding the objective respiratory status of the patients studied. Many reports have provided either limited information (peripheral oxygen saturation, PFTs) or only subjective information regarding the patient’s perceived level of dyspnea. With incomplete information on a minority of patients, we presumed that these patients did not suffer from impaired oxygenation or ventilation. Most of the included studies are case reports, and these reports do not provide consistency across all domains of UPNB administration, such as the type and concentration of local anesthetic or the use of adjuvants. Thus, it is difficult to determine what role, if any, anesthetic type and concentration, as well as the addition of adjuvants, has on the respiratory safety of UPNB. The paucity of literature and heterogeneity in our studies were also limiting factors. The heterogeneity (variability of reported outcomes) hindered pooled analysis of data to perform a meta-analysis. Lastly, given the novelty of the UPNB, there may be an inherent risk of publication bias such that cases of UPNB leading to respiratory complications may not have been published at all.
CONCLUSION
This is the first systematic review demonstrating the respiratory safety of UPNB in spontaneously breathing, appropriately selected patients. Specifically, our findings from the studies show that UPNB does not induce or worsen dyspnea. Besides, UPNB injection produced better saturation, with some studies reporting between 99% and 96-100% oxygen saturation among patients. Patients who exhibited intractable or persistent hiccups had their frequency and severity reduced or vanished after the nerve block injection. Included studies also showed that UPNB preserved ABG samples (fair PaO2, PaCO2, and arterial pH) and spirometry/PFT values (fair FEV1, FVC, and PEF). Presently, the UPNB is rarely used in regional anesthesia practice, but it holds promise for several applications, including pulmonary biopsy, reducing or entirely ceasing intractable hiccups, and lung re-expansion after pneumothorax, with future directions potentially developing. The UPNB carries its merit as an individual nerve block, and it should not be viewed merely as a side effect of the interscalene block, which is associated with its own risks and thus is not justified in cases in which a UPNB would suffice alone. However, the number of patients who have received UPNBs in publications is still small - with large number of studies being case reports and case series - and the quality of those publications is relatively low. Therefore, prospective studies and trials are needed to clarify the true incidence of respiratory complications from UPNBs.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
