A 97-year-old severely demented female on day 1 of data collection was yelling out with movement during care activities or transfer from the bed to wheelchair. She had been in the nursing home 9 months, was nonambulatory, had limited ability to verbally communicate, and her bilateral knee pain was well controlled with scheduled hydrocodone/acetaminophen. She also received scheduled lorazepam for anxiety. Multiple nurse assessments revealed full range of motion to her lower extremities and she denied pain. In addition to yelling, “help me,” “No, No No” or “ooh, ooh ooh,” when approached or moved, she began to be resistive to care and looked very frightened when approached. The staff interpreted this change in behavior as anxiety and responded by approaching care slowly and explaining what was being done and the reason for the care. She was provided positive verbal feedback every time she was “cooperative with cares” and was frequently reassured that “she will be all right and not fall” when transferred. On day 13 of data collection, when asked if she had pain she responded “yes” and touched her upper left leg. She was treated with her scheduled hydrocodone/acetaminophen and positive reinforcement for being cooperative. On day 21, the timing of her scheduled hydrocodone/acetaminophen and lorazepam were changed to 1 h before morning care was delivered, even though she was again denying any pain. On day 22, when asked about pain she pointed to her left leg and when the right leg was touched she yells out. On day 23, an X-ray revealed a left intertrochanteric fracture.

A 91-year-old moderately demented female with a pleasant demeanor, and persistently smiling countenance had been in the nursing home 57 months, was ambulatory, and retained the ability to verbally communicate regarding her back pain. Her back pain was well controlled with scheduled acetaminophen and as needed tramadol. On day 8 she verbally complained about hemorrhoid pain and was treated with medication and a supportive cushion. She presented on day 11 of data collection with a clear change in condition. Though she continued to smile, she became withdrawn, refused meals, and spit out her medications, including her oral analgesics. This behavior continued and she started “spitting out yellow phlegm.” Multiple checks of vital signs were normal, she was afebrile, her lungs were clear, and she had no cough. On day 14 she fell and on day 15 she was restless all night, was spitting up larger amounts of “yellow phlegm.” While continuing to smile, she began to grab staff clothing and jab at them. On day 18, she complained of fatigue, refused to open her mouth and did not complain of pain. A nurse looked into her mouth and multiple “pus pockets” were noted. She was started on an antibiotic. On day 19 she complained of “hurting all over” was very restless and stated “I can’t swallow.” Her daughter ordered hospice care and all scheduled medications were discontinued. On day 22, she was seen by a dentist who diagnosed an acute abscess of the mandibular anterior area that was tender to the touch. She was started on antibiotic and opioid injections, clonazepam orally disintegrating tablets and viscous lidocaine for the jawline. On day 23, she was much weaker and on day 25, she died.

In Case A the resident’s behavioral symptoms were misinterpreted as anxiety and paranoia rather than as the symptom of pain. Caregivers inadvertently treated her anxiety rather than discovering the etiology underlying the behavior. Her history of anxiety probably contributed to this misinterpretation as well as her inability to verbally communicate. Since she had a history of bilateral knee pain, staff focused on physical assessments on the knees and may not have been alarmed by some pain behaviors associated with movement. Since the pain behaviors were present on the first day of data collection it is still unclear how long she had pain from the hip fracture. It was not until day 22 when she clearly pointed to her left leg when asked about pain and yelled out when touched that an X-ray was ordered. This case highlights some of the challenges in weighing the physical discomfort, psychological stress, and costs of diagnostic tests for people with advanced dementia against the potential for identifying and treating conditions that will improve comfort and quality of life.

In Case B, the staff were able to identify an abrupt change in condition. However, pretty clear cues such as stopping eating, refusing medications, and spitting up yellow phlegm were not followed up with a thorough enough assessment. Assessments focused on possible respiratory or urinary tract infections and no one looked into her mouth until 1 week after the symptoms started. This case also highlights how the tendency for older adults to have a blunted fever response complicates the diagnosis of infection. The resident’s pleasant demeanor and persistently smiling countenance may have also contributed to the delayed diagnosis. Research has shown that those residents with more disruptive behaviors such as calling out verbally and physical agitation receive more assessment and treatment than those who are more passive and quiet (Kovach et al. 2006a, b). This case also points to the fact that just because a demented person can consistently verbally communicate some symptom such as back and hemorrhoid pain. Clinicians and caregivers cannot assume that verbal communication will occur for other pain etiologies.

Since behaviors associated with pain in people with dementia may also indicate other physical or psychosocial unmet needs, the use of a differential assessment and treatment protocol can help to identify the person’s unmet need and target treatment more appropriately. The Serial Trial Intervention (STI), developed by our team, was designed for this purpose. The STI is designed to help people in the middle or late stages of dementia. For these people who may be losing their verbal skills and cannot tell you there is a problem, changes in behavior are the only indications that something is wrong. Failure to meet their needs can also have many negative consequences including discomfort for the person, agitated behavior, hospitalizations, resistiveness to care, staff frustration, and death. The STI will be described followed by research results testing use of the STI.

The five steps of the STI are listed in Fig. 10.1. Following identification of behavior change, two levels of targeted assessment and treatment are used. If those steps are not successful in uncovering the problem and successfully treating the person, analgesics and possibly a psychotropic medication are used during Steps 3, 4 and 5. The STI allows assessments and treatments to be customized to the individual’s specific health history and current symptoms and needs. Because many nurses working in nursing homes are practical nurses who have had more limited background education in assessment and pathophysiology, providing extra training to these staff or setting up a system in which a nurse with more education completes assessments is recommended. The main components of our training program are:

We recommend assessing for both pain as well as other changes in behavior using an explicit schedule and procedures for the nurse and ancillary staff to follow. Passive behaviors are often overlooked and may occur when there is pain or another unmet need (Colling 2000). Changes in eating, sleeping, and functional status may indicate pain or other physical or psychosocial problems. We have also found that some people with dementia who have pain will attempt to get out of bed, out of their chair, or off the unit. We have labeled this behavior “exiting” and believe it may be an attempt to get away from their pain. When the person’s pain is treated, the exiting behavior ceases (Kovach et al. 2000). When a resident has a change in behavior that is not effectively treated through basic care provided by the ancillary staff, the STI should be initiated by the nurse. Steps of the STI are outlined in Table 10.2 and Fig. 10.1. Continued movement through steps of the STI is based on the results of assessments and failure of symptoms to improve within a reasonable time frame for that treatment. For example, one would expect to see positive response to an analgesic within a shorter time frame than an antibiotic or an antidepressant. Responses to nonpharmacological comfort interventions such as repositioning, distraction, or massage can be evaluated in shorter time frames. If the symptoms continue after completing all five steps, the process is repeated.