Patient and prescriber perspectives on long-acting injectable (LAI) antipsychotics and analysis of in-office discussion regarding LAI treatment for schizophrenia

Psychiatrists who worked in a community mental health center (CMHC) were selected from a panel previously identified to participate in ethnographic research (ie, a non-random, self-selected sample). Nurse practitioners (NPs), social workers, and therapists were recruited by telephone. All professional participants completed a screening questionnaire to ensure qualification for the study. HCPs provided study information to patients and Health Insurance Portability and Accountability Act (HIPAA) informed consent was acquired from all study participants. HCP participants were compensated for recording their visits with patients and patients were not compensated during this phase of the research.

Psychiatrists and NPs were included if ≥50% of their practice occurred in a CMHC, they treated patients with schizophrenia, and prescribed LAIs for at least some patients. Other HCPs (eg, social workers) were included if >20% of their time was spent in a CMHC and they worked with patients with schizophrenia. Patients in the study had a primary diagnosis of schizophrenia and were indicated for a change in treatment. Patients with a primary diagnosis other than schizophrenia, those who were non-English speaking, or who could not provide informed consent due to cognitive impairment were excluded.

Conversations between patients or their caregivers and HCPs from across the US were recorded during treatment visits between August 2011 and February 2012. Conversations were transcribed and analyzed. Discussions were categorized according to 11 main predetermined topics occurring during a typical CMHC visit for patients with schizophrenia (Figure 1).

A total of 2–4 team members (consisting of 2 medical information researchers and 2 linguists/cultural anthropologists) concurrently analyzed the recorded conversations by reading and listening to each dialogue. Analyses were descriptive and qualitative. Qualitative and linguistic analysis was conducted utilizing a mix of analytical methods including “constant comparison” [31]. Constant comparison is a method that structures analysis of conversational dynamics, lexicon, themes, and meaning across a corpus of data and within cohorts. In constant comparison, the analyst selects 2 data points (here, HCP-patient/caregiver conversations) and analyzes first one side and then the other, for conversational dynamics, lexicon, emergent themes, and essential meanings. Then, the second conversation is analyzed in the same way, noting similarities and differences between the 2 conversations in a spreadsheet and within the texts themselves, such that a rich log of findings and comparative points is constructed and maintained. The analyst then proceeds with a third data point (ie, one side of the next conversation pair), comparing findings from that text to the prior 2 texts and notes any differences or similarities among all 3. Then the analyst compares the fourth data point to findings from analysis of the first 3. This proceeds across the corpus of data within a cohort. Constant comparison is then conducted in other cohorts in this manner. Findings from all cohorts are then compared and contrasted to identify essential differences between and among cohorts.

Further telephone in-depth interviews (TDIs) with psychiatrists (n = 8), patients (n = 12, monthly × 3 months), and caregivers (n = 4) also were conducted to gather follow-up information on LAI discussion, prescription, or use. During TDIs with psychiatrists, audio clips from actual conversations from the patient meeting were played back to gain deeper insight into thought processes during the conversations.

Separately, 4 CMHCs (located in Chicago, IL; Atlanta, GA; Warren, OH; and St. Petersburg, FL) were recruited for in-person observations that took place between December 2011 and February 2012. CMHC psychiatrists provided study information to patients and patients then had to contact study investigators if interested in participating. Informed consent was acquired from all participants in the study. Professional participant and patient/caregiver information collected was made anonymous prior to analysis. The study protocol was approved by the independent New England Investigational Review Board. HCPs were not compensated for their participation, however, the CMHC was provided with a $1500 donation. Patients received a $25 local gift card for participating in the in-person observations.

The research was carried out by teams of 1–2 trained anthropologists. Anthropology researchers spent a few hours at home with some patients prior to their CMHC visit, discussing each patient’s disease and treatment experiences, then followed the patient to the CMHC and observed any meetings with the psychiatrist or nurse, the patient receiving an injection, and any other interactions within the clinic. A total of 15 patient in-home visits, with 10 patients receiving LAIs and 5 receiving oral medications (12 men and 3 women) were conducted.

Prescriber conversations (n = 60 with 14 psychiatrists; n = 9 with 2 NPs) averaged 11.5 minutes in total duration. Duration of interaction varied individually by type of HCP, averaging 12 minutes with psychiatrists, 9 minutes with NPs, and 16.6 minutes with social workers or therapists (4 conversations from 2 social workers and 2 therapists). Conversations between all types of HCPs and patients or caregivers comprised 2 phases: assessment and decision, which each covered approximately 70% and 30% of conversation time, respectively. During the assessment phase of the conversation, patients and/or caregivers dominated the conversation and then generally yielded to HCPs for the decision phase.

Multiple treatment goals were pursued for patients with schizophrenia and were addressed differently by each type of HCP. Social workers and therapists used open-ended questions (“What would you like to talk about?” “What would you like to work on?”) and primarily focused on issues like social wellness and means of achieving daily structure, like work or school. Patients or caregivers sometimes discussed medications and compliance during these sessions but were not explicitly focused on this topic. Prescribers typically used a scripted check-list approach to ensure assessment for positive symptoms, deviating only if positive symptoms were detected and required further investigation. Figure 1 shows a breakdown of the types of topics discussed between prescribers and patients. Treatment discussion and behavior modification/counseling occupied just over 50% of the prescriber-patient visit. Psychiatrists and NPs spent a similar amount of time on treatment discussion while the time focused on behavior modification/counseling is attributed more to psychiatrists (25% vs 1% for NPs). Discussion on compliance occupied only 2% of the prescriber-patient visit. During conversations about medications, prescribers asked simple, direct questions when probing for patient medication compliance (“Are you taking your medications?” “Did you take any medicine last night…?”). Prescribers used direct and logical strategies when probing adherence but could become more authoritative upon discovering noncompliance. Overall conversation flow generally started with prescribers probing for compliance, symptoms, and assessing treatment, which may have included a subjective illness narrative by the patient. The second phase was led by patients where they could express any treatment preferences. The final phase consisted of prescriber-led treatment planning and LAI scheduling (if selected) based on learnings from the earlier phases of the conversation (Figure 2).

A total of 20 unique HCPs (psychiatrists, n = 14; nurse practitioners, n = 2; case/social workers, n = 2; therapists, n = 2) from 16 unique institutions across the United States participated in the study.

Psychiatrists (n = 14) and their patients (n = 60) provided the most complete set of information for the study including recorded conversations, TDIs, and in-person CMHC observations. Psychiatrists’ patients were being treated with oral antipsychotics (n = 22) or LAIs (n = 38). Psychiatrist could be treating individual patients with LAI or oral antipsychotic medications. Psychiatrist and patient characteristics by type of treatment are listed in Table 1. Median years in practice for psychiatrists treating LAI patients and those treating patients on orals were 25 and 18 years, respectively. Patients receiving oral treatment were predominantly female (59%) and LAI patients were mostly male (53%). The majority of patients across both groups (58% to 64%) were initially diagnosed with schizophrenia >10 years prior to the study. More than one fourth (27% to 29%) of the sample were diagnosed within 5 years.

Psychiatrists made antipsychotic treatment decisions without patient or caregiver input during 40 of 60 (67%) conversations. Patients with less severe impairment were more likely to be involved in treatment decisions (conversations with 13 of 36 [36%] mild or moderate patients vs 7 of 24 [29%] severe patients). Involvement in treatment decisions was greater when discussing LAIs: 15 of 60 (25%) with patients/ caregivers vs decisions about oral antipsychotics, 5 of 60 (8%). However, there were no discussions of LAIs by psychiatrists in 11 of 22 (50%) patients taking oral antipsychotics (Table 2), despite the fact that participating patients were indicated for a change in treatment. Overall, only 6 of the 60 conversations (10%) involved patients actively making an antipsychotic treatment decision.

The conversation flow around introducing LAIs typically followed a number of steps that could be terminated by the prescriber or patient at several decision points (Figure 3). More than half (11 of 19 [58%]) of LAI-naïve patients offered LAIs by their psychiatrists agreed to start treatment although just three of those who agreed (3 of 11 [27%]) verbalized favorable responses to an LAI (Table 3). Adherence benefit was the major verbalized reason for accepting an LAI offer and fear of needles was most common for refusals. Almost half of patients offered an LAI were neutral or passive in the decision. During these conversations, a variety of techniques were used to encourage patient acceptance of LAI treatment, including: personal gain to the patient (“…not having to worry about where your pills are…”); sharing other patients’ experience (“Sometimes, patients think that this is easier …”); or occasionally use of fear tactics (“…those voices, those paranoid thoughts are all going to come back. It’s just a matter of time…I can guarantee…that it will happen”). When caregivers were present they were supportive of psychiatrists’ choice of LAIs. However, this was only examined with a small sample, as caregivers were only present in 3 of 19 (16%) discussions with LAI-naïve patients.

If the decision was made to initiate LAI treatment, psychiatrists selected the specific LAI to prescribe with minimal patient input. Only 1 specific LAI was discussed in most of the “new start” conversations (7 of 11 [64%]). Patients and caregivers confirmed in TDIs that LAI selection had been made without their input and they were generally uninformed about choices (“I don’t think there are that many choices with [the] shot.”).

When the discussion about initiating LAIs was abandoned by prescribers, the main reason stated was to preserve a healthy, trusting therapeutic relationship with the patient rather than risk being perceived as coercive. Second, prescribers felt it was important to allow patients to retain autonomy to create treatment “buy-in”. Third, prescribers felt that patients who initially rejected LAIs could become more receptive over time and chose to reintroduce the idea at a later date.

Patient obstacles to LAI use emerged as fear or hesitation about the injections. These perceived fears most consistently impeded LAI prescription choice. Patients who expressed strong concern about injections often did so repeatedly in the conversation (“I never did injections, I don’t like needles…They freak me out, they scare me, they hurt and I don’t like them.”). Despite multiple tactics attempted by psychiatrists, the persistent refusal of patients who expressed a strong fear and concern about injections avoided use of LAIs. Of the 7 of 19 (37%) LAI-naïve patients who responded unfavorably to an LAI offer from a prescriber, only 2 of 7 (29%) received an LAI.

Another barrier to LAI use was the lack of patient insight into the disease and treatment (inability to reason regarding symptoms and treatment options). Patients generalized the negative treatment experience with a single LAI to the entire class of LAIs, even if there was a clear distinction between that past experience and currently available options.

Other examples of potential barriers to LAIs reported by individual patients included the requirement to go to the CMHC to receive injections and wishing treatment with medication was more effective. It is unknown to what extent these reasons ultimately prevented patients from receiving injections. In TDIs, the cost of medication as a barrier was reported in a very small number of cases as most patients received state or federal assistance (8 of 12 [67%]) or held private insurance (3 of 12 [25%]).

Among prescribers, possible side effects were among the chief concerns for LAI usage, specifically, with the long-acting effects of this administration method because treatment cannot be withdrawn rapidly if side effects suddenly occur. Despite this concern, conversations about side effects were rarely initiated by psychiatrists, typically being left for the patient to initiate. Even when specific side effects were explored (eg, with the use of general questions such as, “Have you been sleeping okay?” or “Is your appetite okay?”), they were not always directly attributed to the medication. Similar to psychiatrists, NPs tended to use very general questions about side effects, and during conversations did not always differentiate among side effects of LAIs vs. oral medications. However, LAI treatment changes were rare–only 4 of the 38 patients treated by psychiatrists switched or discontinued LAI treatment: 1 discontinued due to restless legs and other unspecified side effects, while 3 switches to a different LAI occurred due to fatigue/grogginess (n = 1) and high prolactin levels (n = 2). It should also be noted that psychiatrists’ patients on an LAI at time of study had previously received an average of 2.7 injections, suggesting they had probably only recently begun treatment.

Prescribers were most successful in overcoming patient objections to LAIs by decomposing resistance to uncover the severity of resistance and investigate beyond the initially stated problem to address the root issue. Prescribers used several other logic-based approaches to overcome barriers during discussions. Emphasizing the benefits of newer LAIs, like the use of smaller needles with certain injections, or better therapeutic effects with LAIs than their oral counterparts helped patients commence LAI treatment. Empowering patients during the decision (“…just commit to one month of medicine, that’s all, just one shot. If it’s a disaster we’ll switch gears”), emphasizing convenience (“one shot and we can pretty much minimize all medication”), or showing patients the needle and talking to the nurse were also successful approaches. In terms of decomposing resistance, one particular example included a patient who claimed to have a fear of needles, yet was actually resistant due to a 20-lb weight gain with a previous LAI. Digging deeper into the objection was successful. (Patient: “I actually, I have problems with needles. …last year, [my doctor was] giving me those [specific LAI medication] shots and it made me gain 20 pounds in 1 week”, Psychiatrist: “So it wasn’t the injection, per se, it was the side effect of the medicine. There is a different injection we can use. This is a once a month and I have several clients on it who have not gained weight.”).

Twelve patients who received LAIs participated in TDIs; 9 of 12 (75%) believed they had improved over the 3 months during which the interviews occurred and attributed their success mainly to their LAI treatment. No patients reported a worsening condition. Individual patients mentioned benefits of LAIs that included: improvements in symptoms, better concentration, attention, alertness, and a more positive outlook. Patients also mentioned the medication working “faster” and “better” than oral formulations. Generally, strong support systems and an absence of barriers kept patients adherent to the LAI schedule. As with oral medications, stable home environments, involved family members, friends, and other such caregivers, and case workers helped contribute to adherence. Addressing logistical issues like transportation services to enable patients to reach a CMHC for treatment also encouraged LAI usage and adherence.

Information seeking and communication about LAIs may have contributed as another determinant for LAI usage, particularly by using the Internet. Most patients and caregivers reported using the Internet most commonly to search for information about schizophrenia, with most activity occurring around the time of diagnosis. Additionally, a few patients and caregivers reported social media sites (eg, blogs, message boards, chat rooms) to be the most useful source of disease information.

During TDIs, LAI patients reported willingness to share positive experiences with other patients including convenience, efficacy, and concerns about the injections/needles. The above may serve as future assistance for other patients considering LAI treatment.