Hypertensive disorders of pregnancy, including preeclampsia, are commonly associated with headaches. Eclampsia is a form of hypertensive encephalopathy that includes headache, visual disturbances, nausea, vomiting, seizures, stupor, and coma. Seizures may occur in the absence of severe hypertension. A headache is a serious premonitory sign, present in more than 50% of women in whom eclampsia develops. About one-half of the cases of postpartum eclampsia occur within 48 hours after delivery, and the remainder occur between 2 days and 4 weeks after delivery. Occipital or frontal thunderclap headache, blurred vision, scotomas, photophobia, and altered mental status are some of the potential presenting symptoms. Other hypertensive disorders, with or without superimposed preeclampsia, are also associated with headaches both antepartum and postpartum and may lead to encephalopathy.

Posterior reversible (leuko)encephalopathy syndrome (PRES) is a neuroclinical syndrome with specific imaging findings characterized by headaches, seizures, encephalopathy, visual disturbances, and focal neurologic deficits. Conditions associated with PRES include preeclampsia, uremia, hemolytic-uremic syndrome, infection, malignancy, exposure to immunosuppressant drugs, and SARS-CoV-2 infection. Approximately 25% of cases of PRES occur during pregnancy or in the immediate postpartum period. Mayama et al. conducted a retrospective cohort study to assess the incidence of PRES in women with eclampsia and preeclampsia with neurologic symptoms. PRES occurred in 92.3% of women with eclampsia and in 19.2% of women with preeclampsia with neurologic symptoms.

The pathophysiology of PRES is believed to be similar to that of hypertensive encephalopathy in that altered cerebrovascular regulation causes loss of blood-brain barrier integrity. The accompanying vasogenic edema can be reversed by prompt recognition and supportive therapy (e.g., cessation of provocative medications, aggressive treatment of hypertension, seizure prophylaxis) in 70% to 90% of cases. However, irreversible cytotoxic edema with permanent neurologic damage can occur if the initial disorder is not diagnosed early.

The neuroradiologic features of PRES typically include symmetric areas of cerebral edema, predominantly involving the white matter regions of the posterior circulation (occipital lobes, posterior parietal and temporal lobes) ( Fig. 30.1 ). Magnetic resonance imaging (MRI) is the “gold standard” for diagnosing PRES because it can provide information about cerebral involvement earlier than computed tomography (CT).

Posterior Reversible (Leuko)Encephalopathy Syndrome. MR images of a 19-year-old pregnant woman at 31 weeks’ gestation who presented with a urinary tract infection. She developed eclampsia, hemolysis, elevated liver enzymes, low platelet count syndrome, confusion, seizures, and a blood pressure of 190/140 mm Hg. Axial T2 fluid-attenuated inversion recovery images demonstrate cortical/subcortical vasogenic edema in the frontal lobes (the superior frontal sulci), parietal lobes, and occipital lobes as well as deep gray structures such as caudate nuclei and lentiform nuclei (A, B). With magnesium sulfate, blood pressure control, and cesarean delivery, the patient had complete resolution of imaging findings (C, D) and symptoms.

From Brady E, Parikh NS, Navi BB, et al. The imaging spectrum of posterior reversible encephalopathy syndrome: a pictorial review. Clin Imaging . 2018;47:80–89.

The physiologic changes that occur during pregnancy and the postpartum period (e.g., venous stasis, edema, and hypercoagulability) render these patients susceptible to stroke, and a headache is a common presenting symptom. Strokes can be ischemic or hemorrhagic.

The evaluation and treatment of stroke during pregnancy should mimic that performed for nonpregnant patients. Treatment will depend on the etiology. In addition to supportive care, acute reperfusion therapy with fibrinolytic agents (recombinant tissue plasminogen activator) and intra-arterial mechanical thrombectomy should be considered in pregnant women with qualifying strokes.

Ischemic strokes account for approximately 87% of all strokes. Causes of ischemic stroke include cerebral venous sinus thrombosis, preeclampsia/eclampsia, thromboembolism related to valvular heart disease, and profound and persistent hypotension (e.g., cervical arterial dissection and amniotic fluid embolism). The clinical presentation often comprises a new-onset headache that is overshadowed by focal neurologic signs and/or disorders of consciousness.

Hemorrhagic strokes can be subclassified into intracerebral hemorrhage (ICH) or subarachnoid hemorrhage (SAH), with a 10% and 3% incidence, respectively. Hemorrhagic stroke in pregnancy and the postpartum period is relatively more common than in the nonpregnant state. Conditions associated with hemorrhagic stroke include preeclampsia/eclampsia, aneurysms, and arteriovenous malformations.

SAH usually occurs secondary to a ruptured aneurysm or arteriovenous malformation. The classic presentation is a sudden onset of a severe headache that is unlike any previous headache (“worst headache of my life”). Associated symptoms may include neck stiffness, nausea, vomiting, decreased level of consciousness, and focal neurologic deficits. Suspicion of SAH necessitates urgent investigation by CT imaging; nonsurgical therapies (e.g., endovascular ablation) are available, and long-term sequelae can be minimized with early therapy.

Although usually associated with head trauma, subdural hematomas can occur spontaneously during pregnancy or can be associated with dural puncture (see later discussion). In several case reports, identification of the subdural hematoma was preceded by symptoms of PDPH. Dural puncture results in cerebrospinal fluid (CSF) leakage and decreased intracranial pressure (ICP). Presumably, the reduction in ICP causes stress on bridging cerebral vessels, which precipitates tearing and bleeding. Spontaneous subdural hematomas have been reported in parturients with diseases associated with angiopathy, such as preeclampsia and fatty liver disease of pregnancy. Neurologic signs of subdural hematoma are variable but include evidence of increased ICP (e.g., headache, somnolence, vomiting, confusion) and focal abnormalities.

Trager et al. queried a US research network and compared, using propensity matching, the incidence of spontaneous cervical artery dissection (sCeAD) in pregnant and nonpregnant women. They included more than 460,000 women per group; pregnant women had a twofold increased risk of sCeAD (relative risk [RR], 2.06; 95% confidence interval [CI], 1.17 to 3.61; P =.0104) compared to nonpregnant women (8% versus 3.9% in pregnant and nonpregnant women, respectively). Most cases occurred during the pregnancy, not postpartum. Abdelnour et al. performed a systematic review of all published cases of sCeAD in the peripartum period. The identified 77 patients had a mean age of 33.7 years. Risk factors included migraine, hyperlipidemia, connective tissue disorders, preeclampsia and eclampsia, HELLP syndrome (hemolysis, elevated liver enzymes, and low platelet count), and prolonged second stage of labor. The most common symptom was a headache followed by neck pain. The diagnosis was commonly made after carotid vessel ultrasonography or magnetic resonance angiography.

Cerebral venous and sinus thrombosis (CVST) is an uncommon cause of stroke (<1% of all strokes) that results from thrombosis of the cerebral venous system, including dural venous sinuses, and deep and superficial cerebral veins. The most common clinical presentations are signs of intracranial hypertension and parenchymal drainage: headache (70% to 90%), seizure (30% to 40%), papilledema (30% to 60%), focal neurological deficits (30% to 50%), aphasia (15% to 20%), altered level of consciousness (15% to 25%), coma (5% to 15%), and rarely, movement disorder. Diagnosis is best confirmed by MRI in combination with venography. Treatment of cortical vein thrombosis includes anticoagulation and treatment of seizures and increased ICP.

Anticoagulation therapy is recommended for patients with acute CVST, even in selected patients with intracranial hemorrhage. Thrombolysis may be a therapeutic option in a select group of patients with small hemorrhagic infarct and continued neurologic deterioration. Steroid therapy is not recommended.

Intracranial tumors may manifest as a postpartum headache. A headache that is dull rather than throbbing may be an early feature of a brain tumor. Nausea, vomiting, seizures, and/or focal neurologic signs may be present. Neurologic examination may reveal evidence of increased ICP.

Parturients with idiopathic intracranial hypertension (i.e., increased ICP in the absence of a mass lesion, also known as pseudotumor cerebri or benign intracranial hypertension) have headaches and visual disturbances, usually in the antepartum period (see Chapter 47 ). The features of postpartum pseudotumor cerebri mimic the usual chronic headache symptoms experienced by the patient (i.e., nonspecific and varying in type, location, and frequency) and associated symptoms (e.g., visual loss, diplopia, nausea, vomiting, pulsatile tinnitus). The diagnosis is largely one of exclusion. Treatment involves reduction of CSF pressure, either with glucocorticoids, carbonic anhydrase inhibitors, diuretics, or surgical interventions (e.g., optic nerve sheath fenestration, CSF diversion via lumboperitoneal or ventriculoperitoneal shunt, or intracranial venous sinus stenting). Since CSF volume is rapidly replaced, serial lumbar punctures are of limited value and should be reserved for patients who refuse or cannot undergo conventional medical or surgical therapy. Case reports describe the use of an intrathecal catheter for labor analgesia and the administration of an epidural blood patch (EBP) for PDPH in patients with idiopathic intracranial hypertension.

Spontaneous intracranial hypotension is an uncommon cause of a headache that develops following CSF leakage secondary to dural tears. The tears usually occur at the thoracic spinal level and are not associated with prior spinal intervention. An MRI of the brain with contrast will confirm the diagnosis of intracranial hypotension. An MRI of the spine, CT cisternography, or CT myelography with contrast or radioisotope can help identify the exact location of the CSF leak. Presentation of this disorder is identical to that of PDPH because the pathophysiology is the same. The only difference is the lack of a prior neuraxial procedure. Spontaneous intracranial hypotension has been reported during pregnancy and in the postpartum period.

The subdural or subarachnoid injection of air used to identify the epidural space may be associated with the sudden onset of a severe frontotemporal headache, sometimes accompanied by neck pain, back pain, or changes in mental status. A headache is caused by meningeal irritation by air, and symptoms can mimic those of PDPH in that they are worse in the sitting position and may be relieved by lying down. Roderick et al. noted that 2 mL of air injected into the subarachnoid space was sufficient to provoke a symptomatic pneumocephalus. A CT is more sensitive than an MRI to confirm the presence of air within the cranial cavity. The headache typically occurs soon after air enters into the intrathecal space and resolves within 3 to 5 days with reabsorption of the air. Treatment is symptomatic. Administration of oxygen by nasal cannula or face mask may hasten resorption of the air and speed recovery, although this therapy has yet to be proven for pneumocephalus after neuraxial anesthesia.

Meningitis is a complication of neuraxial procedures, and the associated severe headache typically manifests within 12 hours to several days following the procedure (see Chapter 31 ). A headache is accompanied by fever, nuchal rigidity, and the presence of Kernig and Brudzinski signs. Lethargy, confusion, vomiting, seizures, and a rash also may occur. Various strains of Streptococcus , organisms typically found in the upper airway and vagina, have been linked to bacterial meningitis after neuraxial procedures. In several cluster cases, the organisms in the patients’ CSF were matched with the proceduralists’ nasopharyngeal swabs, confirming that these cases of post–dural puncture bacterial meningitis were the result of droplet contamination. Aseptic technique during the neuraxial procedure, including donning of a face mask by the proceduralist, is of paramount importance. The diagnosis of meningitis is confirmed by examination and CSF culture and warrants immediate treatment with antibiotics.

A headache caused by inflamed paranasal sinuses is associated with purulent nasal discharge and, occasionally, fever. Pain may be unilateral or bilateral, depending on the extent of the disease, and the skin over the affected sinus may be tender. The sinuses fill overnight, and the pain typically is worse on awakening. Pain improves in the upright position, which assists drainage.

Caffeine withdrawal may lead to headache, increased fatigue, and anxiety. Caffeine-withdrawal headaches may occur after just 3 days’ exposure to 300 mg/day or 7 days’ exposure to 100 mg/day of caffeine. The average daily dose of caffeine is estimated at 135 mg, which corresponds to 1.5 standard cups of brewed coffee. Although a caffeine-withdrawal headache has not been documented as a cause of postpartum headache, the diagnosis should be considered if the parturient has been drinking caffeinated beverages during the pregnancy.

Askmark and Lundberg reported episodes of intense headache during periods of breastfeeding in a woman known to suffer from migraines. The onset of headaches occurred within the first few minutes of breastfeeding, and the headaches resolved after cessation of nursing. The headaches were associated with an increase in plasma vasopressin concentration. Headaches have also been described in women with breast engorgement who have elected not to breastfeed or have reduced the frequency of breastfeeding.

Sharma and Panda reported a case in which a woman received ondansetron for nausea and vomiting after uneventful spinal anesthesia for cesarean delivery. Several hours later, the patient developed a severe frontal headache that was worse in the upright position and in the morning and evening hours. The symptoms abruptly stopped after the discontinuation of ondansetron. A headache is a common side effect of ondansetron (incidence, 3% to 17%), owing to its antagonism of serotonin 5-HT ₃ receptors, and should be considered in the differential diagnosis of postpartum headache.

Extensive evidence supports the observation that PDPH is uncommon in patients older than 60 years of age and is most common in patients younger than 40 years of age. In older people, the dura may be inelastic and less likely to gape after puncture. CSF leakage may be impeded by adhesions and calcification. The cerebrovascular system also may be less reactive in older patients. Further, this group is less active physically, and older patients may be less likely to complain.

Amorim et al. observed a gender difference with respect to the incidence of PDPH after spinal anesthesia (11.1% female versus 3.6% male). This difference may be related to differences in cerebrovascular reactivity; it is well known that migraine headaches occur predominantly in females and are influenced by hormonal changes. Women may have enhanced vascular reactivity, or perhaps changes in cerebral blood flow are more likely to produce pain in women than in men. A metaanalysis of randomized clinical trials identified a twofold higher risk for PDPH in nonpregnant females than in males.

In a database study of 1,752,243 privately insured patients in the United States, women who had a vaginal delivery following neuraxial labor analgesia had a higher rate of PDPH (0.58%) compared to those who underwent a cesarean delivery following a neuraxial procedure (0.47%). This may be due to pushing during the second stage of labor in women undergoing vaginal delivery. In a retrospective study of 235 parturients who suffered an UDP during epidural catheter placement, the rate of PDPH was lower in women who delivered via cesarean compared with those that delivered vaginally (53% versus 74%). This difference may also be a result of the mechanical consequences of expulsive efforts during the second stage of labor and/or postpartum hormonal changes in cerebrovascular reactivity. Expulsive efforts in the second stage may increase CSF leakage. This possibility has prompted some physicians to restrict maternal pushing after UDP and to use forceps to shorten the second stage of labor. The evidence supporting this practice, however, is conflicting.

Data are conflicting regarding the relationship between obesity and PDPH. In a retrospective study of 518 women, Peralta et al. demonstrated that higher body mass index (BMI) was associated with a lower rate of PDPH. Using a threshold BMI of 31.5 kg/m ² , the incidence of PDPH was 39% and 56% in those with BMI greater than and less than the threshold, respectively. However, pain intensity at headache presentation and the highest reported pain score were similar between high and low BMI groups. In contrast, Song et al., in a retrospective study of 164 women with UDP, did not identify a protective effect of obesity. Possible but unproven explanations of a lower incidence of PDPH in obese patients include increased abdominal pressure, which may reduce the extent of CSF leakage, and/or reduced physical activity in these patients. Other confounding factors, such as differences in the mode of delivery (higher rate of cesarean delivery) and neuraxial opioid administration, may also play a role.

A history of PDPH after previous spinal anesthesia is associated with the development of PDPH with subsequent spinal anesthesia. In a cross-sectional study of nonobstetric patients who underwent spinal anesthesia for elective surgery, those with a previous history of PDPH were 4.3 times more likely (95% CI, 1.99 to 9.31) to have a second PDPH than patients without a history of PDPH (26.4% versus 6.2%, respectively). This finding suggests that certain individuals are predisposed to the development of PDPH.

Seeberger et al. found that multiple dural punctures significantly increased the risk for PDPH. Surgical patients who received a second spinal injection owing to failure of the initial spinal puncture had a 4.2% incidence of PDPH compared with a 1.6% incidence among patients who had a single dural puncture.

Technical factors related to the neuraxial technique influence the incidence of PDPH.

Spinal needle design

Historically, the beveled Quincke needle, a needle with a cutting bevel ( Fig. 30.3 ), was widely used for dural puncture for both diagnostic and anesthetic purposes. In 1951 Hart and Whitacre introduced a solid-tipped, pencil-point spinal needle with a lateral injection port, which is now known as the Whitacre design. They believed that the new needle tip would stretch and separate rather than cut the dural fibers and result in a lower incidence of PDPH. Studies have confirmed the anticipated lower incidence of PDPH with the pencil-point tip (see Table 30.2 ). In vitro data suggest that fluid leak through the dural puncture site is slower with a pencil-point needle compared with a beveled needle. With the recognition that pencil-point spinal needle tips reduce the incidence of PDPH, other tip designs have appeared. In 1987 Sprotte et al. reported experience with the use of a new needle that has a solid oval tip and a longer orifice than the Whitacre needle. The incidence of PDPH was 0.02% with its use in a heterogeneous patient population of almost 35,000 patients. Subsequent studies have shown the incidence of PDPH with the Sprotte needle is lower than that with Quincke needles of smaller gauge. A 2017 metaanalysis of eight randomized clinical trials including 1324 women undergoing cesarean delivery confirmed that spinal needles with cutting bevels were more likely to cause PDPH than pencil-point needles (RR 3.12; 95% CI, 1.60 to 6.10).

Designs of Spinal Needle Tips (Not to Scale).

Illustration by Naveen Nathan, MD, NorthShore University HealthSystem, University of Chicago Pritzker School of Medicine, Evanston, IL.

A modification of the Quincke needle, the Atraucan needle, has a cutting tip and a double bevel, which is intended to cut a small dural hole and then dilate it. Studies suggest that the use of this needle is associated with a higher risk for PDPH than the use of a noncutting, pencil-point needle.

In a 2016 systematic review, Arevalo-Rodriguez et al. reviewed the evidence for prolonged bed rest combined with different body and head positions, as well as supplementary fluid administration, for the prevention of PDPH after dural puncture. The review included 24 trials with 2996 patients; 26% of study participants randomized to bed rest experienced PDPH compared with 21% in women who were allowed to mobilize immediately after dural puncture. Given these data, and the knowledge that pregnant women are hypercoagulable and at increased risk for deep vein thrombosis and pulmonary embolism, immobility should not be used as a maneuver to prevent PDPH (see Chapter 38 ).

Despite the widespread practice of encouraging women to increase oral fluid intake after UDP, there is little evidence that greater hydration prevents PDPH. A 2016 systematic review identified only one randomized trial in 100 nonobstetric patients. There was no difference in the incidence of PDPH in patients randomized to receive 3 L or 1.5 L of fluid per day.

Two clinical trials in nonobstetric patients have evaluated the efficacy of oral or intravenous caffeine to prevent PDPH, but neither study showed a reduction in the incidence of headaches. Based on these results, prophylactic caffeine is not advocated for prevention of PDPH.

Cosyntropin has been studied for the prevention of PDPH after UDP, yielding mixed results. Liu et al. conducted a propensity-matched retrospective analysis of parturients with UDP ( n = 115). Intravenous cosyntropin was administered to 65 patients (55.6%). Among those who received cosyntropin, 37 (57%) developed PDPH compared with 29 patients (58%) in the no-cosyntropin group. EBP was performed in 21 patients (56.8%) who received cosyntropin and 13 patients (61.7%) who did not. These findings were corroborated by Pancaro et al. in another retrospective study in which the benefits of cosyntropin for the prevention and treatment of PDPH after UDP were assessed. The authors did not find a significant difference in the incidence of PDPH between those exposed to cosyntropin (19/32, 59%) and unexposed patients (17/32, 53%; OR 1.37; 95% CI, 0.48 to 3.98; P =.56), or in the incidence of EBP between exposed (12/32, 38%) and unexposed patients (6/32, 19%; OR 2.6; 95% CI, 0.83 to 8.13; P =.095). Both groups of investigators concluded that prophylactic administration of cosyntropin is not associated with a reduced incidence of PDPH.

In contrast to these findings, in an RCT, Hakim found that prophylactic cosyntropin was effective in reducing the incidence of PDPH after UDP in parturients, 33% versus 69%, in the cosyntropin and placebo groups, respectively. The mechanism of cosyntropin is unknown but may be related to an aldosterone-stimulating effect on volume expansion, modulation of pain perception via central endorphin-like action, or increased CSF production by enhanced sodium ion transport.

Adequate dose-response and safety studies have yet to be performed. With the existing evidence, we do not recommend routine use of cosyntropin for the prevention of PDPH after UDP in parturients.

The prophylactic administration of neuraxial hydrophilic or lipophilic opioid for the prevention of PDPH after UDP has yielded conflicting results. In a randomized, blinded trial published in 2008, 50 parturients with UDP and subsequent epidural analgesia were randomized to receive epidural morphine 3 mg or saline-placebo after delivery and again 24 hours later, before removal of the epidural catheter. The incidence of PDPH was 48% in the saline-placebo group and 12% in the morphine group.

In 2020 Peralta et al. published a randomized, double-blind trial that assessed whether prophylactic administration of intrathecal morphine decreases the incidence of PDPH and/or need for EBP after UDP. Parturients who had an intrathecal catheter placed after UDP with a 17-gauge Tuohy needle during intended epidural catheter placement for labor analgesia were enrolled. After delivery, subjects were randomized to receive intrathecal morphine 150 µg or normal saline before the catheter was removed. The incidence of PDPH was 21 of 27 (78%) in the intrathecal morphine group and 27 of 34 (79%) in the intrathecal saline group (difference, −1%; 95% CI, −25% to 24%). There were no differences between groups in the onset, duration, or severity of headache, or presence of cranial nerve symptoms. EBP was administered to 10 of 27 (37%) of subjects in the intrathecal morphine and 11 of 21 (52%) of the intrathecal saline group (difference 15%; 95% CI, −18% to 48%). This study does not support the clinical use of prophylactic intrathecal morphine after an UDP.

Placing a 19- or 20-gauge epidural catheter into the intrathecal space after UDP with an epidural needle has become an increasingly popular technique. The immediate benefits of an intrathecal catheter are continuous labor analgesia using low doses of local anesthetic and opioid, and rapid-onset surgical anesthesia should it be required. Some experts have speculated that the intrathecal catheter might reduce the immediate CSF leak into the epidural space by mechanical obstruction as well as inducing an inflammatory fibrous reaction in the dura rent, thus facilitating closure of the puncture after removal of the catheter. Most studies are retrospective and observational and lack rigorous outcome definitions and follow-up. Data from these studies are conflicting but suggest that intrathecal catheters do not significantly reduce the incidence of PDPH, even if left in place for at least 24 hours after delivery ( Table 30.3 ). Russell conducted a prospective, nonblinded, quasirandomized multicenter study in parturients who had an UDP during epidural needle placement. Participating institutions were randomized in 6-month blocks to either repeating the epidural procedure with placement of an epidural catheter, or placement of a spinal catheter through the dural puncture if UDP occurred during initiation of epidural labor analgesia. Spinal catheters were left in situ for 24 to 36 hours. The incidence of PDPH was not significantly different between women in the epidural (62%) and spinal (72%) catheter groups.

Some studies have evaluated the effect of intrathecal catheter placement on the severity of PDPH. Heesen et al. conducted a metaanalysis that included nine studies of various quality. The authors concluded that placement of an intrathecal catheter did not significantly reduce the incidence of PDPH compared with placement of an epidural catheter. However, they did observe a significant reduction in the need for EBP (RR 0.64; 95% CI, 0.49 to 0.84) after intrathecal catheter placement. These results were reconfirmed in a follow-up metaanalysis 2020. Similar results were reported in another retrospective study that compared laboring patients with UDP who had the epidural resited or had an intrathecal catheter placed. Use of an intrathecal catheter versus resiting the epidural catheter did not decrease the odds of PDPH (aOR 0.91; 95% CI, 0.81 to 1.01), but was associated with a lower need for EBP (aOR 0.82; 95% CI, 0.73 to 0.91; P <.001). The authors found no benefit in leaving the intrathecal catheter for 24 hours for postpartum PDPH (aOR 1.01; 95% CI, 1.00 to 1.02; P =.015) or need for EBP (aOR 1.00; 95% CI, 0.99 to 1.01; P =.40).

The current evidence suggests that advancing a traditional epidural catheter into the intrathecal space after UDP does not appear to reduce the incidence of PDPH but may reduce its severity or the need for an EBP. Leaving the spinal catheter in situ after the patient leaves the birthing unit may increase the risk for drug error or infection. Further studies are required before a definitive conclusion can be reached about the benefits and safety of this technique.

Trivedi et al. randomly assigned patients with UDP to receive either a prophylactic epidural saline bolus (40 to 60 mL) or blood patch (15 mL) just before epidural catheter removal, or conservative therapy without a saline bolus or blood patch. The incidence of PDPH was not different between the saline and control groups (88% versus 67%, respectively). Shah studied 17 patients who received an epidural saline infusion (at a rate of approximately 40 mL/h) for 24 to 36 hours after UDP. Four patients complained of severe intrascapular pain, which resolved when the infusion rate was reduced. Severe PDPH developed in 47% of patients after the infusion was stopped. In contrast, in a nonrandomized, nonblinded study of patients with UDP, Charsley and Abram reported that intrathecal injection of 10 mL of saline immediately before needle or catheter withdrawal resulted in a lower incidence of PDPH (32%) than in a control group who did not receive the saline (62%), as well as a less frequent need for an EBP. In a large retrospective study, Binyamin et al. also found an added benefit of injecting intrathecal saline (10 mL) as it decreased the incidence of PDPH (aOR 0.85; 95% CI, 0.73 to 0.99; P =.04) and the need for EBP (aOR 0.75; 95% CI, 0.64 to 0.87; P <.001). The intrathecal injection of saline after UDP deserves further study. Intrathecal or epidural saline should not be administered until residual local anesthetic effects have resolved.