PA PSRS Patient Saf Advis 2006 Jun;3(2):22-6.
Patient Safety in Mental Health: Reports from Behavioral Health Hospitals
Behavioral Healthcare
Expand All Collapse All

Introduction

Through a Memorandum of Understanding with the Department of Health, behavioral health hospitals (BHHs) licensed by the Department of Public Welfare are required to submit reports to PA-PSRS under the Medical Care Availability and Reduction of Error Act of 2002 (Mcare or Act 13). Several patterns have been identified in over 1,200 BHH reports received during the 18-month period from July 2004 through December 2005. While some patterns reflect the uniqueness of BHHs, there are also similarities with other facility types. Information presented in this article may also be of interest to behavioral health units within other types of facilities.

PA-PSRS information is a foundation upon which facilities can identify opportunities for system improvements. While the following information may provide some insight into BHH report patterns, these patterns may not reflect actual patient safety problems at a specific facility. BHHs may wish to use this information as an impetus to delve more deeply into the causative factors associated with reports in their own facility.

Staff from some BHHs have expressed difficulty identifying reportable events due to the unique aspects of their clinical setting. This article may offer them some insight into the types of events reported by facilities like theirs.

Overall, BHHs submit far fewer reports to PA-PSRS than other hospitals. After adjusting for differences in volume, PA-PSRS receives about seven reports from non-BHH hospitals for every report submitted by a BHH, Even in light of a longer average length of stay, BHHs may have lower report submission rates because their patients tend to have fewer and less serious physical conditions than patients in general acute care hospitals. Further, they typically receive fewer and less risky interventions than patients in other types of hospitals.

As shown in Figure 1, BHHs submit proportionately fewer reports than non-BHHs in every category of event type. For example, Medication Error and ADR report submission rates were approximately one-third of the non-BHH rate. The rate of reports of Falls from BHHs was about one-fifth that from non-BHHs. While this highlights the differences in patient populations and interventions between these types of facilities, it is also worth noting that Medication Errors and Falls are significant patient safety concerns in all of these healthcare settings.

Figure 1. Reports to PA-PSRS by Behavioral Health

 

Figure 2. Distribution of Reports to PA-PSRS by Event Type, Behavioral Health Hospitals (Jul 2004-Dec 2005)

Examining the distribution of reports by Event Type from BHH and non-BHH facilities (see Figure 2) finds some similarities and some differences between these care settings. BHHs have submitted reports for all major Event Type categories except Transfusion. As in non-BHH facilities, Medication Errors and Patient Falls are the most prevalent types of reports.

In contrast, BHHs reported proportionately fewer Errors and Complications Related to Procedure/Treatment/Test; and issues related to Skin Integrity and Equipment. There are a number of reasons for these differences. Patients in BHHs undergo considerably fewer procedures, treatments, and tests compared to patients in non-BHHs. BHH patients are generally more mobile and are therefore at lower risk for skin integrity problems.

While BHHs have reported fewer Medication Errors and Adverse Drug Reactions per patient day than non-BHHs, these types of events constitute a greater proportion of reports from BHHs. In part, this is likely because of BHH patients experiencing fewer opportunities for Errors or Complications Related to Procedures, Treatments, or Tests, but it may also reflect the quantities and/or types of medications administered in these facilities. In addition to maintenance medications for chronic conditions, such as asthma or diabetes, psychotropic medications are associated with extrapyramidal symptoms.

 

Medication Errors

Fifty-five percent (55%) of reports from BHHs were medication errors, compared to 25% from non-BHH facilities. Among both types of facilities’ medication error reports, dose omissions predominated in both. However, BHHs reported a greater proportion of extra dose, wrong drug, and wrong patient errors.

Dose Omissions

Transcription errors contributed to about two-thirds of reports of dose omissions. These involved mainly failure to transcribe, kardex recopying problems, and incorrect/inaccurate transcription. Reported missed doses were often overlooked at bedtime or when the patient was off the unit or out on a pass.

The medication process also broke down at points where orders needed to be clarified or when follow-up was indicated and did not occur. Some of these points include: when an ordered medication was not on formulary/not available; when parental consent was required; or when orders were unclear about the dose/dosing schedule, type/form of medication, timing, or when an automatic stop order was instituted for a necessary medication.

Other instances of reported dose omissions include:

  • Nurses unaware that the medication was in the medication drawer, because the brand name was dispensed instead of the generic name (or vice versa).
  • Parenteral medications not given as ordered when a patient refused oral medications.
  • Computer system problems contributing to omissions.

Extra Doses

Transcribing the wrong medication frequency also caused the majority of extra dose errors. About one-fourth of these reports involved failure to discontinue a medication. Recopying medication kardexes also resulted in such errors. In some cases, medications were administered after the physician discontinued the order. In others, patients received a medication while out on a pass and received an extra dose upon returning to the unit.

Wrong Drug

Wrong drug errors involved several look-alike/sound-alike drugs. For example, the most frequently mentioned wrong drug pair was Klonopin and Ativan. While not mentioned in the reports, these medications have look-alike/sound-alike generic names (Clonazepam, Lorazepam).

Another frequently reported pattern involves confusion between extended release and regular drug versions:

  • Adderall XR versus Adderall (predominant wrong drug error).
  • Lithium Carbonate versus Lithium ER, Lithium SR.
  • Depakote ER versus Depakote.

Patients often received incorrect sound-alike insulin formulations: Humalog versus Humulin, Humulin70/30 and Humalog 75/25. Also, while both are dextroamphetamines, several reports indicated that Adderall was given instead of Dexedrine (and vice versa).

Several wrong drug errors involved other drugs used in behavioral health with similar letter strings. Examples are shown in Table 1. 

Table 1. Similar Letter Strings in Behavioral Health

Almost half of the wrong drug reports indicated the following contributing factors: pharmacy dispensing errors, transcription errors, and misidentification (e.g., wrong patient; wrong kardex referred to; many patients requesting medications at the same time; nurse who administered medications was not the same nurse who poured the medications).

Wrong Patient

In reports involving administration of a drug to the wrong patient, the predominant contributing factors include:

  • Not using two identifiers.
  • Medication was transcribed onto another patient’s medication administration record.
  • Patients with similar names.

In a few instances, instead of asking, “What is your name?” the healthcare worker asked if the patient was a specific name, and the wrong patient said “yes.” In other instances, the wrong patient presented him/herself for the medication when a name was called out, and the healthcare worker assumed that it was the correct patient.

The following systems solutions may help to reduce the incidence of medication errors:

  • Reducing the need for recopying medication administration records.
  • Developing consistent mechanisms to ensure administration of medications at bedtime and when patients return from passes/other departments.
  • Having Pharmacy label dispensed medications with both generic and brand names.
  • Posting a list of brand and generic names of drugs commonly used in behavioral health.
  • Consistently applying a protocol to ensure that ordered parenteral medications are given when patients refuse oral medications.
  • Implementing a system to flag/remind healthcare workers to follow up when medications require clarification.
  • Concerning look-alike, sound-alike and regular vs. extended release medications:
    • Using tall-man lettering system or unique labels to differentiate medications.
    • Separating such medications to avoid confusion.
    • Posting a list of commonly confused medications.
  • Using at least two patient identifiers to ensure that the right drug is given to the right patient.
  • Requiring the person who pours a medication to be the one to give the medication.
  • Pouring and administering one patient’s medications at a time.
  • Moving to Pharmacy-dispensed unit dosing that is administered individually to one patient at a time, rather that prepouring medications for several patients at once.
  • Asking patients to state their names, rather than asking “Are you X?”
  • Regularly sharing medication dispensing error information with the Pharmacy to encourage investigation and process improvement.

Adverse Drug Reactions

Among adverse drug reactions (ADRs) reported to PA-PSRS from BHH, almost two-thirds involved one of the following six medications: Risperdal, Haldol, Seroquel, Lithium, Lamictal, or Trazadone. Three-fourths of the ADRs were: extrapyramidal symptoms and/or rashes.

There were a couple of very serious ADRs. One report revealed that psychotropic medication resulted in neuromalignant syndrome. However, serious reactions were not associated only with psychotropic drugs. In another case, an antibiotic was associated with toxic epidermal necrolysis. Both of these life-threatening conditions required patient transfer for intensive medical intervention.

Methods for preventing ADRs include:

  • Heightening staff awareness of the medications most frequently associated with ADRs, taking into account those used commonly in your facility.
  • Making staff aware of symptoms and interventions, not only of common psychotropic ADRs, but also rare ADRs such as neuromalignant syndrome.
  • Behavioral health clients may also experience ADRs from non-psychotropics, such as antibiotics. Assessment for and intervention in such ADRs may reduce risk in life-threatening circumstances.

Patient Falls

Compared to non-BHHs, a greater proportion of BHHs’ reports of patient falls involved falls while ambulating, in hallways, and on the grounds of the facility. BHH patients are likely to be in better physical condition and are capable of ambulating greater distances than are acute hospital patients. Yet, other factors involved in many BHH falls may be similar to those in acute hospital patients.

The reports indicated several contributing factors that place BHH patients at risk, including: physical symptoms (dizziness, lightheadedness, loss of balance, syncope, seizures); and wet floors. Two-thirds of the falls occurred from 7pm to 7am. Many patients who fell while lying in bed were either asleep or were reaching for something. One-fourth of falls from a seated position involved wheelchairs or benches/chairs tipping.

One-half of the fall reports did not indicate completion of a fall risk assessment or implementation of fall precautions. Yet, two-thirds of these reports specified medications given prior to the fall that may produce symptoms that increase fall risk, such as benzodiazepines, anti-seizure medications, and/or antipsychotics. Of the fall reports specifying medications, 40% indicated that patients received multiple drugs prior to the fall. Polypharmacy also increases fall risk.

Methods to help reduce falls, or to reduce harm from falls, include:

  • Evaluating the circumstances surrounding falls associated with elimination/use of bathrooms.
  • Adjusting medications to reduce dizziness, lightheadedness, loss of balance.
  • Evaluating activities and correcting patient safety hazards in recreational areas, smoking areas, courtyards, sidewalks/walkways, to reduce injuries.
  • Controlling seizures with medication.
  • Securing furniture and equipping ambulatory aids with anti-tip devices, when available.
  • Using partial siderails when patient is asleep and at night.
  • Keeping patient items within reach.
  • Because of medications (particularly benzodiazepines and anti-seizure), assessing all behavioral health patients for fall risk.
  • Implementing fall precautions consistent to level of risk assessed, particularly with patients taking multiple medications, benzodiazepines, and anticonvulsants.

Other Reports

Certain types of reports are more likely to come from BHHs than non-BHH facilities because of the unique clinical setting and patient mix. For example, injuries during recreational activities are frequently reported, as are instances of aggressive behavior in behavioral health patients either towards one’s self or other patients, and possession of contraband. Some reports also indicate cases in which a change in physical condition requires transfer to acute care.

Conclusion

PA-PSRS provides a wealth of information for behavioral health hospitals. PA-PSRS is a springboard from which these facilities can monitor their own data to identify unique patterns and contributing factors. From this information, they can then develop corrective actions, measure outcomes, and ultimately improve patient safety.

PSA LOGO

The Pennsylvania Patient Safety Advisory may be reprinted and distributed without restriction, provided it is printed or distributed in its entirety and without alteration. Individual articles may be reprinted in their entirety and without alteration, provided the source is clearly attributed.

Current and previous issues are available online at http://patientsafety.pa.gov.

©2018 Pennsylvania Patient Safety Authority