Correlates of post-dural puncture headache and efficacy of different treatment options: a monocentric retrospective study

Abstract

Background:

Post-dural puncture headache (PDPH) is a severe positional headache that appears usually within 72 hours after inadvertent dural puncture, secondary to cerebrospinal fluid leakage. It is treated first by conservative treatment (including bed rest, hydration, caffeine and simple analgesia) and then by invasive procedures such as blood patch. This study aims to evaluate factors associated with PDPH among a sample of Lebanese patients and assess the rate of success of different treatment modalities administered in a specific sequence: conservative treatment first, then ultrasound-guided bilateral greater occipital nerve block (GONB) if failure of conservative treatment and finally epidural blood patch (EBP) if failure of GONB.

Methods:

A retrospective case–control study was conducted between January 2015 and December 2019 in the Notre-Dame des Secours University Hospital. Out of a total of 10,051 procedures, 18 cases were diagnosed with PDPH and were matched based on gender, age and procedure type to a control group of patients who did not develop PDPH randomly selected (72 patients).

Results:

Higher body mass index (adjusted odds ratio (ORa) = 0.77) was significantly associated with lower odds of PDPH, whereas the presence of previous chronic headache (ORa = 5.56) was significantly associated with higher odds of PDPH. Seven out of 18 (38.89%) had their pain symptoms/headache resolved on conservative treatment. Out of the remaining 11 patients, 6 (33.33%) had their symptoms resolved on GONB with a significant decrease in the pain score 48 hours after GONB compared to baseline (5.55 vs 9.73; p = 0.007). Five patients (27.78%) had their symptoms resolved using a blood patch, with a significant decrease in the pain score after blood patch compared to baseline (3.00 vs 9.80; p = 0.041).

Conclusion:

Our preliminary data suggest that ultrasound-guided GONB is a minimally risky and efficacious technique for those who fail to respond to conservative treatment.

Background

PDPH (Post-Dural Puncture Headache) is a frontal–occipital dull severe headache, resulting from cerebrospinal fluid (CSF) leakage and consequential intracranial hypotension. It is exacerbated in standing or sitting positions and relieved in the supine position. PDPH happens after either inadvertent dural puncture during epidural anaesthesia or procedures that involve intentional dural punctures such as spinal anaesthesia and diagnostic lumbar puncture. ¹ PDPH is, therefore, a significant cause of lawsuits filed against anaesthesiologists in the United States ² and one of the causes of patients’ refusal and dissatisfaction regarding neuraxial anaesthesia. ³

The overall incidence of PDPH varies from 0.1% to 36%, with the highest incidence usually seen with diagnostic lumbar puncture using standard Quincke cutting needles of 20 or 22 Gauges. ⁴ Associated symptoms are common and may include neck stiffness, nausea, vomiting, dizziness, visual disturbances such as photophobia or diplopia, and occasionally auditory symptoms such as hearing decrement and tinnitus. Presence of fever, focal neurological signs, papilledema or a headache unrelieved by pain medication warrants elimination of life-threatening diagnoses such as meningitis. ⁵ Most PDPH commonly occurs within 72 hours of dural puncture and after up to 2 weeks in some rare cases. PDPH can resolve within 1 week; however, it might persist and requires specific treatments such as epidural blood patch (EBP) or saline injection. ⁶

Diagnosis of PDPH relies usually upon the clinical features of the headache and the history of dural puncture and if in doubt imaging such as magnetic resonance imaging (MRI) helps to establish the diagnosis by showing meningeal enhancement or extradural CSF collection. ⁷ PDPH is the result of multiple risk factors including the following:

1. Female gender. ⁶

2. Age (between 20 and 50 years old). ⁸

3. Needle properties such as size (the bigger the size, the higher the risk for PDPH), ⁹ design (Quincke cutting needles) ¹⁰ and bevel orientation (a perpendicular orientation to the longitudinal axis of the dural cylinder). ¹¹

4. Operator inexperience. ¹²

5. Obstetric perinatal pathologies. ¹²

6. Individual comorbidities such as low body mass index (BMI) ¹³ and previous chronic headaches. ¹⁴

Still, there is some controversy if migraine can be considered a risk factor for PDPH with one prospective study showing that migraineurs are not at increased risk of developing PDPH; ¹⁵ however, Haller et al. ¹² have considered migraine as a patient-related risk factor for PDPH in their root-cause analysis. Depending on these risk factors, the incidence of PDPH varies widely from one institution to another. ⁴

First-line treatment of PDPH is conservative. It consists of bed rest, hydration (oral or intravenous) and simple analgesia, defined as peripheral acting drugs that are effective by mouth (such as acetaminophen and non-steroidal anti-inflammatory drugs). ¹⁶ Bed rest has a temporary effect only and should be accompanied by appropriate prophylaxis for unwanted thromboembolic events. Caffeine has a short-term effect and its administration should be limited to 24 hours. Cases of maternal seizures were described for large doses of caffeine (more than 1 g per day); thus, three cups of coffee (300 mg) are recommended per day in the setting of PDPH. Stronger opioids if used should be limited to 72 hours because of their side effects. ¹⁷ Wagner et al. ¹⁸ have reported an approximated early efficacy of conservative treatment of 30% in patients with PDPH. It is considered that most cases of PDPH (85%) resolve spontaneously within 6 weeks and that the conservative treatment has mostly a temporary pain relief effect.^17,19

EBP is viewed by many as the gold standard for the treatment of PDPH. ⁹ It is a procedure in which autologous blood is injected at the spinal tap site to seal the dural leak by a simple coagulation process. Some postulate also that it elevates the CSF pressure by compressing the Dura and that its effect is not different from that of normal saline injection. Yet, this technique can have severe rare complications such as bleeding, epidural abscess, cranial nerve palsy and nerve compression, which can lead to neurologic damage. The procedure by itself can be ineffective or partially effective as some patients require sometimes a second blood patch. Side effects such as back pain, neck pain and fever as well as central nervous system (CNS) infection (meningitis) are common after EBP. It is important to mention that EBP is mostly contraindicated in cases of coagulopathy and systemic infections.^20,21 The success rate of EBP is between 61% and 98%, with a lower rate of success in the obstetric population, related to the use of large epidural needles. ²²

Greater occipital nerve block (GONB) has been used in clinical practice for many years to treat complex headache syndromes refractory to medical treatment such as occipital neuralgia and chronic migraine. For GONB, a linear array probe of 8–13 MHz is required and scanning should begin at the midline 3–4 cm below the external occipital protuberance and then moved laterally if needed.^23,24 Ultrasound guidance for performing the GONB is considered reliable and safe as patients are at least two times less likely to encounter multiple skin puncture or vascular puncture.^25,26 Walker et al. ²⁶ had concluded in the Cochrane systematic review that the use of ultrasonography improves the success rate of peripheral nerve block procedures and results in better onset time.

In the literature review, publications such as case reports and observational studies have shown some efficacy of GONB in the treatment of PDPH. ¹⁷ One randomized study done by Naja et al. ²⁷ has compared nerve stimulator-guided occipital nerve blockade to conservative treatment (bed rest, hydration and simple analgesia) and demonstrated a statistically significant reduction in the visual analogue scale (VAS) for pain. Akyol et al. ²⁸ have used the ultrasound-guided GONB for PDPH and demonstrated high efficacy of about 57% (in a sample of 21 patients). Unlike the EBP, ultrasound-guided GONB is a very safe technique. There is some evidence to suggest the efficacy of GONB in the treatment of chronic headaches including PDPH. However, guidelines for the management of PDPH still recommend the use of EBP after the failure of conservative treatment because more evidence regarding the effectiveness of GONB is required.

This study aims to evaluate factors associated with PDPH among a sample of Lebanese patients and assess the rate of success of different treatment modalities administered in a specific sequence: conservative treatment first, then ultrasound-guided bilateral GONB if failure of conservative treatment and finally EBP if failure of the GONB technique.

Methods

Ethical aspect

The ethics committee at the Notre-Dame des Secours University Hospital (CHUNDS) hospital approved the study protocol.

Study design

A retrospective case–control study was conducted between January 2015 and December 2019 in the Notre-Dame des Secours University Hospital (CHUNDS) University Hospital.

Contraindications to EBP (presence of coagulopathy or active infection) were among the exclusion criteria. Exclusion criteria were therefore as follows: (1) age less than 18 years old, (2) not able to be contacted, (3) refusal to participate in the study, (4) having a coagulopathy, (5) having a systemic infection and (6) not treated by the following sequence: conservative treatment first, then ultrasound-guided bilateral GONB if failure of conservative treatment and finally EBP if failure of the GONB technique.

All records of procedures with a potential risk of PDPH including spinal, epidural analgesia and lumbar puncture between January 2015 and December 2019 were reviewed. Out of a total of 10,051 procedures, 18 patients were diagnosed with PDPH following the procedure described in Figure 1.

Figure 1.

Flow chart showing cases of PDPH per procedure type.

Data collection was done using a combination of medical records and oral structured questionnaires (through phone calls). Oral questionnaires were double-checked with medical records for recall-bias, to verify answers and strengthen data reliability.

Collected data consist of characteristics related to non-modifiable risk factors for PDPH (such as age, gender, BMI, history of chronic headache or migraine before the procedure) and experienced symptoms along with the onset of PDPH (such as nausea, vomiting, neck pain, dizziness, photophobia, diplopia, tinnitus and hearing decrement). Other characteristics target conservative treatment (bed rest, caffeine consumption, simple analgesia and opiates use). Medical records were also checked for information related to spinal/epidural needle type (Quincke cutting needles, pencil point needles or epidural Tuohy needles) and size (Gauge); both considered as modifiable risk factors for PDPH.

In our observational study, ultrasound-guided GONB was done 1–2 weeks after the failure of the initial conservative treatment and EBP was done 3 days after GONB when lacking clinical improvement. EBP and ultrasound-guided GONB were performed by the same operator. Numeric Pain Rating Scales (NPRSs) were obtained before and 48 hours after each procedure (GONB and EBP) from medical records and physician’s follow-ups. For further assessment, closed-ended questions (yes/no format) were employed to determine if each treatment modality was a failure or success in the patient perspective. It should be noted that managing PDPH with ultrasound-guided GONB following failure of conservative treatment had already been a part of the hospital protocol and was not solely added for the purpose of conducting this study.

These 18 patients who developed PDPH were matched by gender, age and procedure type to a group of control patients who did not develop PDPH or any known dural injury by checking medical records. Randomization was done in a second time for eligible controls (case:control ratio of 1:4). Information regarding BMI, previous chronic headache and migraine, before the procedure and onset of symptoms after the procedure, as well as procedure’s characteristics was similarly collected for controls to cases.

Ultrasound-guided bilateral GONB has been done in 11 (all were females and aged between 25 and 44 years old) out of the 18 patients 1–2 weeks after the failure of the conservative treatment (which includes bed rest, hydration, caffeine, simple analgesia and opiates in some cases). The GONB technique was done using 5 mL bupivacaine and dexamethasone. A blood patch has only been done on the third day after the failure of the GONB technique to alleviate the headache. All patients reported having PDPH were eligible to participate in the study, as no one had met any exclusion criteria.

Statistical analysis

Data analysis was done using SPSS version 23 software. The chi-square test was used to compare two groups. Multivariable logistic regression was conducted using the presence versus absence of PDPH as the dependent variable and taking all variables that showed a p < 0.05 in the bivariate analysis as independent variables. In all cases, p < 0.05 was considered statistically significant.

Results

The mean age of all patients was 32.12 ± 8.49 years (88.9% females). A significantly higher mean BMI was found in controls compared to cases (24.78 vs 22.82). Comparisons in terms of sociodemographic and other characteristics of the patients are summarized in Table 1. Among cases, 6 (33.3%) patients were bedridden during a part of the day, whereas 12 (66.7%) were bedridden all the time and 5 (27.8%) have received opioids. All cases (100%) have received simple analgesics and caffeine.

Table 1.

Sociodemographic and other characteristics of the patients (N = 90).

Variable	Controls (N = 72)	Cases (N = 18)	p
Gender			1
Male	8 (11.1%)	2 (11.1%)
Female	64 (88.9%)	16 (88.9%)
Pregnancy			1
No	44 (68.8%)	11 (68.8%)
Yes	20 (31.3%)	5 (31.3%)
Type of procedure			1
Caesarean section	24 (33.3%)	6 (33.3%)
Vaginal delivery	20 (27.8%)	5 (27.8%)
Other gynaecologic procedure	4 (5.6%)	1 (5.6%)
Urologic	4 (5.6%)	1 (5.6%)
Orthopaedic	4 (5.6%)	1 (5.6%)
Lumbar puncture	16 (22.2%)	4 (22.2%)
Age (in years)	32.13 ± 8.50	32.13 ± 8.50	1
Body mass index (kg/m2)	24.78 ± 2.79	22.82 ± 2.35	0.006

Bivariate analysis

A significantly higher number of cases had a history of previous chronic headache (33.3% vs 8.3%), and developed neck pain (27.8% vs 0%) and dizziness (16.7% vs 0%) compared to controls (Table 2).

Table 2.

Bivariate analysis of factors associated with headache among the sample.

Variable	Controls (N = 72)	Cases (N = 18)	p
Needle size			0.351
19	20 (27.8%)	5 (27.8%)
20	3 (4.2%)	0 (0%)
22	13 (18.1%)	4 (22.2%)
25	3 (4.2%)	3 (16.7%)
27	33 (45.8%)	6 (33.3%)
Needle tip type			1
Pencil point	36 (50.0%)	9 (50.0%)
Quincke	16 (22.2%)	4 (22.2%)
Epidural Tuohy	20 (27.8%)	5 (27.8%)
Previous chronic headache			0.012
No	66 (91.7%)	12 (66.7%)
Yes	6 (8.3%)	6 (33.3%)
Previous migraine episodes			1
No	64 (88.9%)	16 (88.9%)
Yes	8 (11.1%)	2 (11.1%)
Nausea			0.284
No	63 (87.5%)	14 (77.8%)
Yes	9 (12.5%)	4 (22.2%)
Vomiting			0.195
No	67 (93.1%)	15 (83.3%)
Yes	5 (6.9%)	3 (16.7%)
Neck pain			<0.001
No	72 (100.0%)	13 (72.2%)
Yes	0 (0%)	5 (27.8%)
Photophobia			0.200
No	72 (100.0%)	17 (94.4%)
Yes	0 (0%)	1 (5.6%)
Dizziness			0.007
No	72 (100.0%)	15 (83.3%)
Yes	0 (0%)	3 (16.7%)

Numbers in bold indicate significant p-values.

Multivariable analysis

The results of a forward logistic regression, taking the presence versus absence of PDPH as the dependent variable, showed that higher BMI (ORa = 0.77) was significantly associated with lower odds of PDPH, whereas the presence of previous chronic headache (ORa = 5.56) was significantly associated with higher odds of PDPH (Table 3).

Table 3.

Multivariable analysis: forward logistic regression taking the presence versus absence of PDPH as the dependent variable.

Variable	p	Adjusted odds ratio (ORa)	95% Confidence interval
Body mass index	0.013	0.77	0.62	0.95
Previous chronic headache (yes vs no a )	0.014	5.56	1.41	21.90

PDPH: post-dural puncture headache.

^aReference group.

Resolution of the pain symptoms/headache in cases

Seven out of 18 (38.89%) had their pain symptoms/headache resolved on conservative treatment. Out of the remaining 11 patients, 6 (33.33%) had their symptoms resolved on GONB with a significant decrease in the pain score 48 hours after GONB compared to baseline (5.55 vs 9.73; p = 0.007). Five patients (27.78%) had their symptoms resolved using a blood patch, with a significant decrease in the pain score after blood patch compared to baseline (3.00 vs 9.80; p = 0.041). No significant difference was found in terms of experienced symptoms between those who benefitted from the conservative treatment and those who did not and between those who benefitted from the GONB and those who did not (p > 0.05 for all associations).

Discussion

This case–control study shows that patients with PDPH had significantly lower BMI compared to controls, in agreement with previous studies. Previous chronic headaches, as many suggest, may be considered also an associated factor with PDPH. No difference in terms of previous migraine episodes is seen between cases and controls. Neck pain and dizziness were among the most specific symptoms attributed to PDPH in our cases, along with the headache. This study has examined the efficacy of ultrasound-guided bilateral GONB in patients who failed to respond to conservative treatment and shows a success rate of 55% for this technique. The rates of success for the early conservative treatment (which includes bed rest, simple analgesics and caffeine intake) and for EBP were in agreement with other previous studies.

Incidence

Out of 10,051 patients who underwent procedures with a potential risk of PDPH (receiving either spinal anaesthesia or epidural lumbar punctures or lumbar punctures) in our retrospective monocentric study between January 2015 and December 2019, 18 patients were reported experiencing PDPH. The incidence of PDPH among all procedures was 1.79 cases per 1000 over 5 years (or 0.358 cases per 1000 per year). The highest incidence of PDPH in this study was seen with lumbar puncture procedures (1.29 cases per 1000 per year). In the literature, PDPH incidence varies widely from one study to the next. In this circumstance, Jabbari et al. ⁴ have stated that the overall incidence of PDPH after dural puncture varies from 0.1% to 36%, depending on multiple risk factors including patient risk factors (e.g. gender, age and BMI) and procedural risk factors (e.g. needle type, needle size and bevel orientation for cutting needles). However, PDPH depends also on operator experience.

Factors associated with PDPH

The mean BMI in our findings is 22.82 and significantly higher in controls (24.78). Many publications suggested an inverse relationship between BMI and incidence of PDPH; thus, patients with high BMI are less likely to get PDPH after unintentional dural puncture. ²⁹ In that perspective, Faure et al. ²⁹ reported an increased incidence of PDPH in patients with BMI <30 kg/m² compared to patients with BMI ⩾30 kg/m². However, not all studies have demonstrated a relationship between a high BMI and lower incidence of PDPH: Webb et al. ³⁰ reported that parturients who developed PDPH after 17 Gauge Tuohy needle had higher BMI compared to those that did not develop PDPH. There is a known association between the fact of pushing a lot during the delivery and development of PDPH, ³¹ and it is yet unknown if any interaction exists between BMI and pushing during delivery.

The study shows that 33.3% of cases had significantly a previous history of previous chronic headache compared to controls (8.3%) in agreement with what Clark et al. ¹⁴ have stated that previous chronic headache is a presumed risk factor for PDPH. No difference in terms of previous migraine episodes was seen between cases and controls, in agreement with what Van Oosterhout et al. ¹⁵ have demonstrated that migraineurs are not at increased risk of developing PDPH.

The female:male ratio in our reported cases is 8/1, and this ratio is congruent with other studies that showed an increased risk of PDPH in females.^6,32 Factors that might explain the high incidence of PDPH in women vary between anatomical, physiological and socio-behavioural. Experiences were done on human cadaveric Dura mater using a system of dural sac and acrylic column mimicking the in-vivo scenario of dural puncture. After perforating the cadaveric Dura, the liquid outflow was greater in female cadaveric Dura than in male-derived parts. This can be accounted for a thinner Dura in women, which facilitates cerebrospinal fluid leakage. ³³ The high level of oestrogen in women plays also a crucial role in excessive cerebral vasodilatation in response to intracranial hypotension. ⁶ In addition to that, females are usually more sensitive to pain than men, having a greater activation of the prefrontal cortex, insula and thalamus in response to nociceptive information. ³⁴

The mean age in our sample was 32.11 years, which is consistent with other studies that demonstrated an increased risk of PDPH in young people.^6,8 Three factors can explain why PDPH is more prevalent in adults rather than in the elderly. Premenopausal women aged below 50 years have higher cerebral vasodilatation response to CSF pressure drop by the mean of high oestrogen levels. ³² Besides, through aging, the Dura mater loses its elasticity due to atherosclerosis and mechanical changes, making it more difficult for CSF to leak from the dural hole. Moreover, the vertebral extradural space diminishes with age, limiting the CSF accumulation and creating an extradural resistance enough to stops CSF leakage. ⁶

Experienced symptoms

Along with the hallmark postural characteristic severe headache, the bivariate analysis showed that a significant number of cases had neck pain and dizziness, compared to controls. Loss of CSF results in low intracranial pressure and decreased brain buoyancy causing a caudal shift of intracranial structures. The effect of this motion will be a stretching of the Dura and traction on intracranial pain-sensitive structures. Besides traction on meningeal vessels and dural folds that produce the classical headache, it is important to mention the possibility of traction on cranial nerves which explain the many associated symptoms. Nausea and vomiting are induced by the activation of chemoreceptors in the medulla oblongata after traction on the vagus nerve. Neck pain may result from tension on cervical nerves. ³⁵ Photophobia comes from trigeminal nerve disturbances given that eye structures such as conjunctiva, cornea, sclera, uvea, and orbital blood vessels are innervated by trigeminal fibres. ³⁶ Given the communication between the subarachnoid space and the inner ear via the perilymphatic duct which makes the perilymph continuous with CSF, any change in the intracranial pressure will affect middle ear functions. For instance, any decrease in the intracranial pressure following CSF leakage can result in vestibular disturbances leading to dizziness. ³⁷

Efficacy of conservative treatment

Early conservative treatment was efficacious in 38.89% of cases. This result is in agreement with the findings of Wagner et al. ¹⁸ (early conservative treatment efficacy of 30%).

Efficacy of occipital nerve block and blood patch

Out of the 18 reported cases of PDPH, 11 underwent ultrasound-guided GONB after the failure of conservative treatment. Six patients out of 11 had their symptoms resolved on GONB with a significant decrease in the pain score 48 hours after GONB compared to baseline. A blood patch was performed after 3 days in five patients who did not respond to GONB. Therefore, the efficacy of ultrasound-guided bilateral GONB is about 55% in this study. When administered after the failure of GONB, EBP was efficacious in all cases. This result is in agreement with what has been stated by Kwak ²² (success rate between 61% and 98%).

Role of ultrasound-guided occipital nerve block in the resolution of pain

Greater occipital nerve is formed by sensory fibres from C2 and C3, and it innerves the medial posterior scalp up to the anterior vertex. The greater occipital nerve usually lies in the fascia between the obliquus capitis inferior and the semispinalis capitis, medial to the greater occipital artery. ²³

GONB results in pain inhibition over the sensorial area, by interrupting the afferent signal to the dorsal horn and the trigeminal nucleus caudalis (TNC) neurons and thereby ‘winding-down’ central sensitization.^23,24 Before the use of ultrasound in neuraxial anaesthesia, GONB was done using either landmark technique (done by identifying greater occipital artery pulsations) or nerve stimulator technique, but these approaches lack efficacy due to the variable anatomical course of the nerve. ²⁴

The idea of the usage of occipital nerve block for the treatment of PDPH comes from Blumenfeld et al. ³⁸ review of peripheral occipital nerve block indications for primary headache disorders, secondary headache disorders and cranial neuralgias. Moreover, according to the International Classification of Headache Disorders (ICHD-3), PDPH is a secondary headache attributed to low cerebrospinal fluid pressure. ³⁹

In the literature, some studies are assessing GONB efficacy using either ultrasonography, nerve stimulator or the landmark technique for the treatment of PDPH.^{24,27,28,40–42} The variable course of occipital nerve can lead to unsatisfactory results when using the nerve stimulator technique. Ultrasonography has increased the success rate of GONB and reduced complications and side effects. ²⁴ In a retrospective analysis done by Bomberg et al., ²⁵ the use of ultrasound for nerve block has shown to reduce the odds of vascular and multiple skin punctures compared to nerve stimulator technique. However, local anaesthesia systemic toxicity (LAST), a life-threatening adverse effect that occurs when local anaesthetics reach systemic circulation, was not observed in any patient with ultrasound guidance technique alone, but it was seen in one patient with the combined use of ultrasound and nerve stimulator and in one patient with nerve stimulator technique alone. However, the use of ultrasound increased the odds of unintended paresthesias. ²⁵

Limitations

As we are dealing with small sample size, we were not able to assess the effects of needle size and type on PDPH onset. The sample was taken from one hospital; thus, the results cannot be generalized. The operator experience is considered one of the principal factors associated with PDPH, but it was not analysed. Mild cases of PDPH may not be reported in medical records, so the incidence of PDPH in our monocentric study might be underestimated. It is a retrospective study with low evidence level; to better evaluate the efficacy of different treatment modalities, including the new ultrasound nerve block technique, prospective studies such as randomized controlled trials are recommended. Finally, NPRS is a subjective measure, and therefore, it consists of one of this study’s limitations.

Conclusion

Our preliminary data suggest that ultrasound-guided GONB is a minimally risky and efficacious technique for those who fail to respond to conservative treatment. Therefore, following the sequence of managing PDPH with conservative treatment first, then GONB and finally EBP for those who remain symptomatic can be considered for optimizing the quality of care. A randomized controlled trial for better evaluation of the efficacy of ultrasound-guided GONB for the treatment of PDPH is needed.