Cardiopulmonary Compromise Associated With Milk and Molasses Enema Use in Children

Objective: Constipation is a common childhood problem that frequently confronts pediatric practitioners. When enema treatment is indicated to treat constipation, milk and molasses enemas are sometimes used. Side effects and adverse outcomes have been described with many types of enemas, but not with milk and molasses. The purpose of this study is to describe five children who acutely decompensated after receiving milk and molasses enemas.

Cited Here...: Retrospective review of the medical records of children with constipation who received milk and molasses enema with subsequent PICU admission at Cincinnati Children's Hospital Medical Center between 1994 and 2000.

Cited Here...: Five children developed significant hemodynamic deterioration after receiving milk and molasses enemas. One of these children died and the others recovered after aggressive resuscitation.

Conclusions: Milk and molasses enema administration can be associated with significant hemodynamic compromise. These risks appear similar to those risk observed with the use of other enema preparations.

INTRODUCTION

Enemas are used in many settings and for many purposes. These purposes include bowel cleansing before radiologic procedures, aiding in the administration of food and medicine, and treating constipation and encopresis in children. Enema preparations often used in pediatrics, such as sodium phosphate (Fleets, Fletchers), barium, soapsuds, and saline, have been associated with infrequent but important complications. Multiple reports describe metabolic derangements in serum phosphate, magnesium, sodium, calcium and potassium with the use of sodium phosphate, saline and magnesium phosphate enemas (^1-15). Cardiac arrhythmias have also been reported, most frequently in adults, during barium enema administration (^16,17). Transient bacteremia has been noted with the use of barium, air, and coffee enemas (^18-22). Colonic perforation has been reported with over-the-counter enemas, barium, and saline enemas, and colitis has also been associated with the use of soapsuds enemas (^23-25). The reported colitis was of short duration. Finally, allergic and anaphylactoid reactions have been observed when saline, barium, and herbal preparations have been used as enemas (^26,27). There have been no reports describing complications or adverse effects with the use of milk and molasses enemas.

In our institution, milk and molasses enemas have been routinely used when enema therapy was indicated because of our belief that this enema preparation had minimal risk associated with its administration. First described in a 78-year-old nursing text, these enemas were used for carminative purposes (to relieve flatus) (²⁸). They were said to work because the sugar in the molasses is irritating to the intestinal lining, and the milk and sugar together produce gas which distends the intestines, causes pressure, peristalsis and evacuation. The following case studies detail several adverse events that occurred in our institution following the use of milk and molasses enemas in children with serious medical problems.

MATERIALS AND METHODS

The cases were identified by searching the records of all children who were admitted to the Intensive Care Unit at Cincinnati Children's Hospital Medical Center between January 1994 and December 2000 and who had constipation (ICD-9 code 564.0) as one of their discharge diagnoses. This search identified 140 patient records. These records were reviewed to determine if a milk and molasses enema had been administered and if any potential adverse effects from the enema administration had prompted admission to the PICU. Data from 43 admissions could not be located. Five cases wherein adverse effects from enema use had prompted PICU admission were identified. Data describing the history, physical examination, laboratory values, resuscitation history, and outcome were extracted from the records, analyzed, and are reported here.

RESULTS

Case 1

C.R. was a 3-year-old male with multiple medical problems including chromosome 2p duplication, panhypopituitarism, bronchopulmonary dysplasia, Tetralogy of Fallot, and seizures. He was admitted to the hospital for increased seizure activity associated with new onset apnea and bradycardia. He also was noted to have abdominal fullness at the time of the initial physical examination felt to be secondary to constipation. His routine medications included Dilantin, Klonopin, Nitrofurantoin, Zinc, Claritin, Neo-Calglucon, Iron, Synthroid, Cortef, Poly-vi-sol, Albuterol, and Flovent. Upon admission, he was placed on his home regimen of CPAP of 10 cm H₂O via tracheostomy. The increased seizure activity was treated with an intravenous bolus of Dilantin and maintenance therapy was initiated. Constipation was treated with two milk and molasses enemas (2 ounces whole milk mixed with 2 ounces molasses). Shortly thereafter the patient developed hypoxemia with oxygen saturations of 70% to 80%, increased abdominal distention, hypotension with blood pressures of 55/34 torr and bradycardia to 91 beats/min. Within minutes he became asystolic requiring a full cardiopulmonary resuscitation including fluid boluses, epinephrine, dopamine, and dobutamine infusions. After return of spontaneous circulation the patient was noted to be anuric. Laboratory values just before the arrest were notable for hyperglycemia (270 mg/dL), normal calcium, magnesium, phosphate and electrolytes, bicarbonate of 21 mmoL/L, and elevated transaminases (SGOT 9151 IU/L, SGPT 9566 IU/L). A blood gas obtained at the beginning of the arrest revealed a pH of 7.19, pCO₂ of 49 torr, PaO₂ of 56 torr, and -9.6 base deficit. His white blood cell count had increased to 25,900/mm³ compared to 11,500/mm³ at the time of hospital admission. His prothrombin time was 34.7 sec and his partial thromboplastin time was 53.9 sec. An abdominal radiograph revealed anasarca and ascites without free air. Surgery was consulted and diagnosed the patient with abdominal compartment syndrome. The hypoxemia and acidosis continued to deteriorate despite aggressive resuscitation. His parents chose to forego surgical decompression, support was withdrawn, and the child died.

Case 2

T.F. was a 20-month-old white female with vesicoureteral reflux, a history of imperforate anus, vesicostomy, small ventricular septal defect, failure to thrive secondary to food intolerance, and constipation. She was admitted to the hospital to receive intravenous antibiotics to treat an Enterococcus urinary tract infection that had failed outpatient management. Before admission, she had developed vomiting and was unable to tolerate her medication. She had also become more sleepy and listless and had a 400-gram weight loss. On initial abdominal exam, large palpable masses of stools were noted in all quadrants. Abdominal radiograph showed large quantities of stool. A milk and molasses enema (3 ounces whole milk added to 3 ounces molasses) was administered. Within 10 minutes, the child became mottled, pale, tachycardic with heart rate 190 beats/min, blood pressure 97/40 torr, distended abdomen, waxing and waning mental status, and weak peripheral pulses. She was transferred to the PICU where she underwent tracheal intubation and mechanical ventilation. She received 80 mL/kg of normal saline, and inotropic support. Initial venous blood gas revealed pH 6.80, PaCO₂ 13 torr, PaO₂ 84 torr, base deficit -24.5. Electrolytes, calcium, magnesium, and phosphate were normal except for hyperglycemia (194 mg/dL). Complete blood count was normal. Urinalysis revealed ketones, occasional white blood cells, no leukocyte esterase or nitrites, and trace bacteria. Her prothrombin time was 15.6 seconds, partial thromboplastin time 57.8 seconds, fibrinogen 50 mg/dL, and D-Dimer>1. Following resuscitation in the PICU she underwent manual stool disimpaction. She was supported in the PICU with 3 days of mechanical ventilation, inotropic medications for 48 hours, blood products, and broad-spectrum antibiotics. All urine and blood cultures were negative during her hospital stay. She recovered without significant morbidity and was discharged home 10 days after hospital admission.

Case 3

W.C. was a 7-month-old white male with past medical history significant for anterior meningomyelocele repaired at three weeks of age. He had a history of rectal stenosis and chronic constipation that had been managed by anal dilatation and Lactulose therapy. During the few days before admission, decreased stool volume was noted and his family sought care for possible stool impaction. On the day of admission, his primary medical doctor identified a mass in the child's left lower quadrant that was consistent with stool. He recommended a milk and molasses enema. An enema (1 ounce whole milk added to 1 ounce molasses) was administered at home and immediately thereafter the patient developed large volume loose stool output that lasted for approximately one and one half hours, along with three episodes of emesis. The child then became lifeless and was brought to the Emergency Department. At the time of Emergency Department admission, the child was gray in color, poorly perfused, with an initial heart rate of 172 beats/min, blood pressure that was unobtainable and temperature 100.8 °F. Intravenous fluid resuscitation was initiated, laboratory studies were obtained, and antibiotics were begun. Notable laboratory values included glucose 298 mg/dL (after receiving dextrose), white blood cell count 17,200/mm³, and capillary blood gas pH 7.23, PaCO₂ 41 torr, PaO₂ 91 torr, and base deficit -11.6. Electrolytes, calcium, magnesium, and phosphate were normal. Abdominal and chest radiographs were unremarkable. Following resuscitation, his mental status improved. A review of systems revealed no history of fevers, recent illnesses, or sick contacts. At the time of admission to the PICU, his temperature was 99 °F, pulse 164 beats/min, respiratory rate 33/min, and blood pressure 101/33 torr. He was awake, quiet and non-interactive. His color was pale, and his abdominal exam was benign. Rehydration was continued overnight and he began to eat again the next day. He slowly returned to baseline mental status. Blood, urine, and stool cultures were negative for bacterial pathogens and rotavirus. No white blood cells were found in the stool. He received ceftriaxone for 3 days during his hospital stay and was discharged home after 3 days.

Case 4

M.A. was a 4-year-old white female with a history of metatrophic dwarfism, seizures, hydrocephalus, ventriculoperitoneal shunt, gastroesophageal reflux, gastrostomy tube, and constipation. Her home medications were Zantac, Propulsid, and Depakote. During a visit to her primary care clinic for immunizations, hard stool was noted in her left lower quadrant. A milk and molasses enema (35 mL whole milk with 35 mL molasses) was administered in the clinic and she immediately developed pallor and hypotension with BP 70/40 torr and pulse 150 beats/min. She was transferred to the emergency department where she received 30 mL/kg nor mal saline boluses. Her blood pressure improved to 100-120/ 40-70 torr and pulse 120-140 beats/min. She received manual disimpaction in the Emergency Department by the surgical service. Abdominal radiograph showed bowel filled with stool, no dilated loops, no free air, and no air-fluid levels. At the time of admission to the PICU, her pulse was 100 beats/min and blood pressure 113/40 torr. An abdominal exam revealed a baseball-sized stool mass in the left lower quadrant. The abdomen was tender to palpation and diminished bowel sounds were noted. Her perfusion was good. Laboratory studies were notable for sodium 141 mmol/L, potassium 2.7 mmol/L, chloride 111 mmol/L, bicarbonate 19 mmol/L, Glucose 172 mg/dL, Magnesium 3.1 mg/dL, and phosphorus 2.7 mg/dL. Calcium was normal. Venous blood gas was pH 7.23, PaCO₂ 44 torr, PaO₂ 94 torr, and base deficit -9.3. White blood count was 18,700/mm³. Over the course of her PICU stay, she required an additional 15 mL/kg of normal saline; however, no inotropic support was required. She was supported with maintenance intravenous fluids. The stool impaction eventually resolved with the manual disimpaction, Go-lytely, and Dulcolax suppositories. The patient was discharged in stable condition.

Case 5

A.H. was a 6-year-old male with a chromosome 22 abnormality, asthma, and recurrent pneumonias, who was admitted for decreased responsiveness, perioral cyanosis, and cool extremities occurring 1 hour after the administration of a milk and molasses enema (1 ounce whole milk with 1 ounce molasses). He had been in his usual state of health on the day of admission but had not had a bowel movement in 4 days. His normal home medication was Lactulose. He had a large bowel movement after the enema and 3 bouts of emesis just before the acute changes noted above. On examination in our Emergency Department, his pulse was 112 beats/min, respiratory rate 24/min, temperature 100.4 °F, and blood pressure 83/71 torr. He responded only to painful stimuli. His abdomen was tender to palpation and no masses were palpable. His electrolytes, calcium, magnesium, and phosphate were normal except for bicarbonate of 16 mmol/L and a BUN of 22 mg/dL. White blood count was 4,700/ mm³ (57% neutrophils, 18% bands, 21% lymphocytes). Arterial blood gas was pH 7.38, PaCO₂ 31 torr, PaO₂ 156 torr, and base deficit -7.8. Abdominal radiograph revealed stool and gas throughout the bowel without free air. Urinalysis, lumbar puncture, and liver function tests were all normal. Air contrast enema was negative. He received minimal fluid resuscitation in the Emergency Department and was admitted to the general pediatric floor for further rehydration. He was treated with ceftriaxone for possible sepsis. All cultures (blood, urine and CSF) were negative and he was discharged to home in his usual state of health.

DISCUSSION

These cases identify patients who developed acute respiratory and/or hemodynamic decompensation soon after receiving a milk and molasses enema. Each patient had serious underlying medical conditions, and all but one had chronic constipation. The underlying medical status of four of the children put them at higher risk for cardiopulmonary compromise compared with otherwise healthy children.

There are several possibilities that might account for the acute decompensation after enema administration though no single explanation accounts for the circumstances in all five cases. Clinical deterioration associated with electrolyte abnormalities have been previously reported with the use of many types of enemas. These abnormalities include hypokalemia, hypocalcemia, hypophosphatemia, acidosis, and hypernatremia (^3-10,12,13). Hypermagnesemia with fatal or near-fatal outcomes has been reported following magnesium sulfate enemas used in neonates for hyaline membrane disease (^1,15). Hyponatremia with central pontine myelinosis has been described with tap water enema use (²). Moore et al. (¹⁴) described children with an enema syndrome where hyperkalemia was noted following administration of a variety of different types of enemas. In their study, hyperkalemia was a predictor of fatal outcome. We did not find any significant electrolyte disturbances in our patients, with the exception of one patient with mild hypokalemia.

Another risk factor for fatal outcome and clinical deterioration, identified by Moore et al., was leukocytosis (¹⁴). In our patients, leukocytosis was observed along with an increased percentage of band forms, and low-grade fever. This laboratory finding may reflect an infectious etiology such as transient bacteremia or a stress response associated with enema administration. Transient bacteremia has been noted following enema administration and may result from either direct mucosal injury or increased intraluminal pressure and bacterial translocation (^18-22). LeFrock et al. found that bacteremia associated with barium enema use was of immediate onset and brief duration (^19,20). These authors suggested that the transient nature of the bacteremia was due to bacterial clearance by the liver and that patients with compromised hepatic function might be at higher risk of persistent bacteremia (²¹). Transient bacteremia was a possible cause of shock in our patients, although we did not document any patients with positive blood cultures.

Rectal perforation is another described hazard of enema administration and is suggested to occur from either tip injury or increased intraluminal pressure (^23-25). We did not observe any radiographic findings consistent with bowel perforation in our patients and it is unlikely that this mechanism could account for the cardiopulmonary compromise.

Another possible mechanism for cardiopulmonary compromise following enema administration is a cardiac arrhythmia (^16,17). Eastwood found that 46% of patients undergoing barium enema had electrocardiographic changes. These changes were more common in the elderly and in those patients with underlying heart disease (¹⁶). Two of our patients had underlying heart disease and were therefore at risk for cardiac arrhythmias. Electrocardiograms were not obtained following enema administration in our patients, but it is possible that cardiopulmonary dysfunction may have been preceded by an abnormal cardiac rhythm.

Allergic reactions have been observed with barium enemas, herbal additives, and plastic enema tips (^11,26,27). It is possible that an allergic reaction to either the milk or molasses was responsible for the physiologic changes noted in our patients.

Abdominal compartment syndrome is a possible explanation for one of our cases (²⁹). The cardinal feature of this syndrome is an elevated intra-abdominal pressure, secondary to peritoneal fluid surrounding the viscera, significant swelling of the viscera, or a large amount of luminal contents within the viscera. With extremely high intra-abdominal pressures, venous return via the inferior vena cava is impaired and cardiac output is attenuated. Increased abdominal pressure on the diaphragm often results in compromised ventilation. In most of our patients, the absence of organ swelling, ascites, or a significant amount of stool does not seem to fit this explanation. It is possible that the hypertonic enema resulted in intestinal wall edema that might have contributed to increased intra-abdominal pressure.

The pathophysiologic mechanism for the cardiopulmonary compromise observed in our patients remains unclear. Gut bacterial translocation and transient bacteremia resulting from increased intracolonic pressure seems to be the most likely etiology. Another possible explanation is that significant fluid shifts associated with the osmotic effects of the milk and molasses could have led to increased stool output and hypovolemic shock.

The North American Society for Pediatric Gastroenterology and Nutrition has developed clinical practice guidelines for constipation (³⁰). Oral or rectal medications are recommended for functional constipation with fecal impaction in children one year of age and older. Osmotic enemas should be avoided in children less than 2 years of age. Enemas have been associated with injury to the rectal wall, abdominal distention and emesis, and appropriate precautions should be taken to avoid these complications. The Society further recommends consultation with a Pediatric Gastroenterologist if attempts at disimpaction are unsuccessful. No recommendations have been made regarding milk and molasses enema preparations.

The cases presented in this article indicate that enema administration has risk, even using seemingly benign preparations such as milk and molasses. Children with chronic illnesses are at a higher risk for constipation and may also be more likely to suffer adverse outcomes with the use of enemas. When enema therapy is unavoidable, low-volume enemas with short retention times may reduce the risk of complications. Intravascular volume status should be evaluated before using enemas involving hypertonic solutions of any kind and cardiac health should be considered. Enemas should be administered by individuals who are well trained to manage potential complications associated with this therapy.