Hemolysis: Crisis Intervention
Elisabeth Harman

Hemolysis remains a concern for all workers in dialysis facilities. In spite of the many CQI factors we put into place to check on the safety of our equipment and procedures, there are still times when circumstances beyond our control intervene to give us a wake-up call. This report outlines events that occurred in our dialysis facility: the symptoms we encountered, our response, and the outcomes of the three patients who were hospitalized.

Program Description
Our clinic is a 22-chair freestanding, hospital-owned facility in a western state. We serve 63 patients on chronic hemodialysis 6 days/week and also have a Home PD program.

Case #1. On Saturday January 15, 2005, we were operational as usual with one exception. We were dialyzing a visiting patient from one of our sister units who was receiving plasmapheresis in preparation for a mis-matched living donor transplant. HP is a 30-year-old patient without diabetes who was on dialysis following a transplant rejection. Her usual facility is a small rural one that is only open 3 days per week and, since she was receiving plasmapheresis on her regular dialysis days, we dialyzed her in our facility. We have an MIQS information system in place in all of our facilities and it is, therefore, very easy to stay current with all patients in all facilities. We used the same dialyzer that she had been using in her home facility and, as they are a non-reuse facility, we used a dry pack for her. Our protocol with dry packs is a 1000 cc normal saline flush and then recirculation with another 500 cc. Her access was cannulated and treatment was started. Within about 10 minutes, she began with pruritis, which worsened. The on-call physician was notified and IV Benadryl 25 mg was ordered with a repeat times one if needed. The assumption was first use syndrome. After the initial dose of Benadryl, her symptoms subsided only a little and the second dose was given with good effect. The itching settled down and she went to sleep. During her run, it was noted on her blood volume monitor that she was trending positive and her hematocrit was dropping. This looked a little unusual and she was monitored very closely. Her vital signs remained stable and she tolerated treatment. At the end of treatment, her needles were pulled and she was up and about talking to others and had no complaints. She was discharged in good condition.

The following day, Sunday January 16th, HP went into the hospital for a plasmapheresis procedure and, at that time, she complained of abdominal pain. When blood was taken for her procedure, she was found to have very significant hemolysis and she was sent from the lab to the ER where she was admitted. Over the course of her hospitalization, which was about 5 days, she refused blood transfusions because of her transplant status. With conservative care and monitoring, she was discharged on January 20th and she returned to dialyze in her home facility on January 21. The cause of her hemolysis was unknown at that time, and when she was admitted the nephrologist on-call spoke at length to our staff to try to determine what he felt was a dialysis-induced hemolysis, but could also not come up with a cause.

Case #2. CG was another patient in our facility who dialyzed on the same T-TH-Sat schedule as HP. He is a 54-year- old patient with diabetes and numerous co-morbid conditions, including having had a seven-bypass heart surgery. He has been dialyzing for just over 10 years.
On January 20, 2005, CG came in for treatment at his usual time. He is on reuse and has been for many years. He requested that his target weight for this treatment be 74 kg. He weighed 78.9 kg pre-treatment. He had a headache pre-treatment, which is usual for him, and he was given two Tylenol tablets, which he also gets every treatment.

Sterilant residual was checked and signed off by two staff. Water checks had also been done. CG’s access was cannulated and treatment started, at which point, he took the Tylenol. He asked to have his chair reclined to a Trendelenberg position so he could sleep. He was routinely checked and it was noted that his blood volume monitor that had begun to trend down at the beginning of treatment was now starting to trend up. His blood pressure also climbed up. His initial systolic BP was 128/——, and by mid-treatment had climbed to 160/170. It was felt that his position in the chair with his feet elevated made his vascular refilling occur more quickly, which would account for the increasing positive result on the blood volume monitor. When the blood volume continued to increase, the concern was that perhaps the monitor was inaccurate. It was turned off and restarted. When it was resumed, the graph began to indicate a negative result.

With about 30 minutes left in the treatment, CG awoke and complained of feeling cold. He had some shaking as if with chills of a fever, but his temperature was 97 degrees. He was asked if he had ever had this shaking with dialysis before and he indicated that he did when it is cold outside. It was noted that his face was flushed, his BP was still elevated, and his heart rate had doubled from pre-treatment rate. He complained of nausea and had some dry heaves with a small amount of white frothy mucous. We had removed approximately 5 kg of fluid at this time and it was felt that perhaps we had removed too much fluid. His target was decreased to minimum and he was watched very closely.

With 15 minutes left in his treatment, his temperature spiked to 99.3 degrees and the nurse practitioner was immediately notified. Blood cultures were drawn and he was ordered to receive 2 gms of Ancef after treatment. After retransfusion, CG seemed slower to respond to questions, though he remained oriented and gave appropriate answers. He was slightly diaphoretic on his forehead. His systolic BP remained elevated and he was given 500 cc of NS due to the possibility that too much fluid had been removed. He stated that he felt somewhat better following this. He was then given the Ancef. He remained slow to respond and flushed. He also continued to have dry heaves, but they were not severe or episodic. The nurse practitioner was called again and ordered an additional 500 cc of NS.
 
He did not improve after receiving the NS, had another episode of nausea, and expectorated frothy white sputum, which had a small red streak in it. He remained flushed and, though appropriate, seemed more lethargic. Emergency services (911) was called and he was transported to the hospital. He was assessed in the hospital, and a lab specimen was found to be grossly hemolyzed. He was transferred to a trauma hospital and admitted to ICU. He complained of severe abdominal pain in the hospital.
Over the course of his hospitalization, which was 6 days in duration, he was monitored in the ICU. He was stabilized and received 2 units of blood when his Hct fell low enough that cardiology felt he should be transfused. His Hct rose to the low 30s and remained there for the rest of his hospitalization.

At this point, the etiology of his hemolysis was unclear and, after his discharge, he dialyzed at one of our sister facilities until the problem was resolved.

Case #3. MB dialyzed in our facility on the same days as the above two patients: Tuesday, Thursday, Saturday. He dialyzed as well on Thursday, January 20, and had no problems during his run. He ultrafiltrated appropriately and had no adverse reactions while in the unit.

The following day, MB presented to the ER with complaints of abdominal pain and nausea. He was admitted to a med-surg floor and monitored. The presumptive diagnosis was pancreatitis, but when his blood was drawn, he was found to have a moderate hemolysis. He was treated conservatively and discharged in about 2 days.

Discussion
With this third event, it was decided that for the safety of all patients the unit should close until a cause could be established. Our patients were accommodated at our sister facility. Our unit has a mixture of machines, but all are manufactured by the same company. The company was notified immediately of the problem and, in an effort to help solve the issue, they emergency shipped us a new lot of bloodlines and then proceeded with their part of the investigation.

All machine logs, water logs, and results of testing were reviewed. Water was tested from every conceivable port in the building, including regular tap water from the kitchen. All these results came back negative for any contaminants. Even the tap water from the kitchen had a chlorine level that was considered safe for dialysis by AAMI standards. As we methodically eliminated the usual causes of hemolysis, the focus seemed to narrow down to bloodlines. There was no kinking of the bloodlines on any of these patients during their runs. We drew blood samples on all of our patients, including the ones at our sister units to compare the lactose dehydrogenase levels on all of them. There were elevations noted in a significant number of patients, who at this point were still asymptomatic. The only common denominator amongst the units was the bloodlines.

We had a conference with a physician researcher from another dialysis facility in the city. After looking at all of our data, including the symptoms and hospital courses of the three patients who were admitted, he came to the conclusion that the bloodlines were the likely source of the problems. However, he did not want to conclusively say this until the company providing the bloodlines could finish their investigation. During this investigation period, our CQI coordinator was instrumental in gathering and organizing the reams of data collected. She participated in and kept notes during all meetings with the company, the outside investigators, and the administrators from our hospital. She was responsible for the chain of events log that was kept and for the “Adverse Events Checklist” that is now part of our policy and procedures.

The company was extremely helpful and proceeded with their own investigation. All of our bloodlines that we had been using were recalled to include several sets of used lines from the day in question. These were dissected and tested at their base of operations and they were unable to find any narrowing or other cause that could be directly attributed to the bloodlines. In the meantime, we were supplied with new tubing. Although we have not been able to conclusively say that the bloodlines were the issue, the problem has been resolved with the new lines we are using.
Our unit was closed for 2 days while this investigation and testing was performed. We dialyzed our M-W-Fri patients on Sunday in our sister unit and our T-Th-Sat patients on Tuesday as they had fewer patients and could accommodate ours. We sent our own staff to work with our patients in this other unit. We re-opened on Wednesday January 26th and, over the next 2 weeks, lab samples were sent for lactose dehydrogenase and hemoglobin and hematocrit on all patients.

Our patients were kept informed during this process and appreciated that we made arrangements for their safety and well being during this time. Not one of them complained at having to dialyze in another facility when the concerns were made clear to them. Our staff responded with exceptional flexibility and kindness and concern for our patients.

Summary
Hemolysis is something that occurs to a small degree with every dialysis treatment. As we monitor our patients and the equipment we use, we hope that we have looked at everything and that our patients will be safe. We were extremely fortunate that we did not have a mortality related to this event. All of the hospitalized patients recovered and HP went on to receive her kidney transplant and is doing well. CG came back to dialyze in our unit after his physician was satisfied that our problem had been solved.

In retrospect, we did not realize that we had an instrument in use that picked up the hemolysis long before we did. Our blood volume monitors showed us the increase in blood volume as the cells were lysed and vascular volume increased as the intracellular fluids were released. The concurrent rise in BP with these patients was also indicative of increased vascular volume. The blood volume monitor also showed the drop in hematocrit as red cells were lysed. We are now more experienced in interpreting these parameters and we are grateful to have this monitor on all of our patients and to have a new parameter to watch for to help forestall any further problems. Our patients’ symptoms were baffling at the time because there were no overt signs of hemolysis. There was no “cherry pop” or brown colored blood in the lines or dialyzers which are what we have been taught to look for. Patients #1 and #3 essentially had no symptoms till the next day. Patient #2 exhibited nausea and some mild abdominal pain, which was also not new for him with his diabetic gastroparesis.
 
Our nursing staff was very “gun shy” following this episode and some staff members seriously considered a change in profession. Fortunately they did not and opted to stay. It was brought home to all that dialysis, though it may seem to be a routine procedure, is inherently risky. Diligence and careful monitoring of all of our patients during their runs needs to be our utmost priority. Hemolysis to this degree is fortunately rare and we are grateful to have had an education with a positive outcome for all concerned.

Hidden Hemolysis
Elisabeth Harman

Ms. X is a 72-year-old female on chronic hemodialysis for 39 months at our 34-station outpatient dialysis center. She presented for her second dialysis treatment of the week on the third patient shift of the day.

Initial assessment: BP: 125/75, P: 70, Pre Wgt.: 55.1 Kg, Dry Wgt.: 52.5 Kg.

Presenting Complaints: Patient noted chronic back pain and mild shortness of breath. Medical History: Coronary artery disease, S/P myocardial Infarction, AAA repair, hyperlipidemia, hyperparathyroidism, thoracic aneurysm, chronic abdominal pain. Allergy: PCN.

Routine monthly lab work was drawn predialysis and treatment was initiated. Ms. X had an uneventful dialysis until the last 40 minutes when she developed moderately severe abdominal pain and worsened back pain. Her vital signs at this time were a BP of 141/84 and a pulse of 87. The treatment was stopped 13 minutes early. Although the patient experienced some improvement, she refused the recommendation for evaluation in the emergency room and instead went home. Two hours after arriving home, Ms. X again began feeling unwell and was evaluated by her neighbor, who is an intensive care nurse. At this time, the nurse demanded that Ms. X go to the emergency room and the patient agreed.

ED Assessment: Chief complaint – Abdominal pain. Physical exam was unremarkable. Abdominal CT scan documented a large aortic abdominal aneurysm. No evidence of either rupture or dissection of the aorta.
TCV Consultation: The aneurysm was not believed to be the cause of the abdominal pain.

ED Lab work: Hgb.: 7.2 g/dl. The lab noted gross hemolysis. First specimens able to be read noted an elevated LDH greater than 2200 (norm 100-250) and Troponin level of 1.5 (normal is less than 0.6). The lab supervisor remarked that the serum of these specimens was black and that in his “thirty years of working in the lab he had never seen serum that looked like this – not even in a cadaver.” Other differential diagnoses were considered and ruled out. The reality that this most likely resulted from something about the hemodialysis treatment was becoming clear. Upon review of the routine predialysis blood work that had been drawn hours before, a Hgb of 13.8 and no hemolysis were noted, but shockingly the routine postdialysis kinetic modeling specimen was hemolyzed.

Now the gravity of the situation was clear and intense scrutiny and investigation began. Immediately, the hemodialysis machine used was removed from the treatment area and called in for service evaluation. It was noted that before this machine was identified and sequestered, three patients had been dialyzed with it and none had developed any symptoms or medical complications.

Treatment and Outcome: Ms. X required three (3) units of packed red blood cells for her 5.8 g/dl drop in hemoglobin. The hemolysis ended within the first 24 – 48 hours after hospitalization although other medical complications occurred. Ms. X recovered and was discharged from the hospital to rehabilitation after 27 days with no sequelae.

The Investigation Continues
Although the ordeal was over for this patient, the etiology remained elusive. A final quality and root cause analysis multidisciplinary group was formed to bring together information and organize the investigation. Literature reviews were conducted to guide the process. As the hemolysis was not evident at the time of treatment, all disposables from Ms. X’s equipment had been discarded and were not available for analysis. However, all similar lots of dialyzers and blood lines were taken out of stock. As this incident was formally reported to all involved manufacturers, we notified them of any immediate resupply required for sequestered supplies.
The point that needs to be made here is that all elements, supplies, and processes utilized to perform the therapy must be evaluated. These include:

1. Water treatment system, including cleaning and testing processes, water analysis and cultures.
2. Hemodialysis machine;
3. Disposable supplies; report incident to all involved manufacturers.
   a. Recommendation: save the dialysis circuit of any unusual occurrences or suspected hemolysis for    investigational purposes.
   
4. Hemodialysis Treatment Record – review of arterial and venous pressures, dialysate composition, temperature and pressures.
5. Dialysis processes – cleaning and residual testing of equipment with documentation, daily quality testing of all equipment. Keep an open mind. Evaluating all procedures may bring to light changes that have occurred over years that need to be revisited as areas for improvement.
6. Enhanced patient monitoring may be deemed necessary to insure this is not a widespread problem. This can range from identifying key signs and symptoms to routine blood sampling as surveillance to test for hemolysis during the treatment.
7. In-service all staff to review signs and symptoms of hemolysis and any enhanced monitoring. We must go back to basics and remember that things do not always present in a textbook manner. Hemolysis is not necessarily visible as “cherry red blood” as we were always taught!
8. Inform all patients of any necessary information as they are active participants in their care.
9. Report incidents to appropriate local regulatory agencies as well as nationally via the FDA voluntary Med Watch reporting program. Working together as a team can help to identify the cause and assess if this is an isolated or recurring issue. Get the word out – you may not be alone!