Factors Affecting Use of Insulin Pens by Patients With Type 2 Diabetes

Data were obtained through computer-assisted telephone interviews of panels from a national sample of people in the U.S. self-identified as having diabetes. Sampling and data collection was performed by GFK, a contract research organization. An institutional review board approved this study.

Potential respondents were contacted by telephone and asked to participate in a survey about diabetes. Inclusion criteria for participants in the current study were: 1) self-reported diagnosis of diabetes by a physician (≥1 year before survey), 2) age ≥35 years (thus, almost all participants had type 2 diabetes), 3) taking insulin ≥1 year, and 4) had heard of insulin pens. Patients who had ever used an insulin pump were excluded, as were those who had participated in diabetes-related research in the previous 3 months. The responses of 600 patients are reported here, including 300 who were taking insulin using a vial and syringe and 300 who were using a pen.

Data collected from participants (Table 1) included: 1) basic demographic information; 2) disease characteristics; 3) medical insurance coverage for pens; 4) whether pen use was encouraged by the patient's physician, discouraged by the physician, discussed as an option, or not discussed at all; and 5) participants' perceptions of how pen use compared with vial and syringe use (6-point scale with pen much better = higher score and pen much worse = lower score) in terms of clinical effectiveness (controlling blood glucose levels), convenience, cost, and facilitation of self-care (making it easier to adhere to the diabetes self-care regimen). Patient questionnaires were constructed on the basis of focus groups conducted with patients and physicians. Hierarchical stepwise logistic regression was used to identify factors that independently discriminated between those who were currently using a pen and those who had never used a pen.

Table 1 shows the results of logistic regression analyses conducted to discriminate those who used a pen (scored 1) from those who did not (scored 0). The first analysis (column 2, labeled step 1) forced into the model a number of control variables. The second analysis (column 3, labeled step 2) added physician action measures that met the forward stepwise criterion as independent variables (P < 0.05), i.e., whether the physician encouraged pen use and whether the physician mentioned pen use as an option. The third analysis (column 4, labeled step 3) added measures of patients' pen perception that met the forward stepwise criterion as independent variables (P < 0.05), i.e., respondents' perceptions of higher pen effectiveness in improving self-care and lower pen cost. Variables that did not meet the forward stepwise criterion (i.e., perceptions of pen clinical effectiveness and convenience) were not entered in the latter analysis.

Physician recommendation of pen use powerfully discriminated pen users from nonusers (odds ratio 135.6). Other factors that discriminated pen users included physicians' presenting pens as an option (14.1) and patient perceptions that pens facilitate diabetes self-care (20.2) and are not costly (4.8).

Patients with type 2 diabetes do not depend on exogenous insulin for survival, but over time most of these individuals will require supplemental insulin, as a consequence of decreased endogenous insulin production (9–10). Yet many patients with type 2 diabetes do not initiate insulin therapy as soon as they should (11–13). Alternative insulin delivery systems such as pens could lower some barriers to initiating insulin therapy, but little is known about factors associated with patients' use of pens. This study considers these issues in a national panel sample of adults with type 2 diabetes who were taking insulin and had heard of pens.

Physicians' presentation of insulin pens was the most powerful predictor of pen use. Merely presenting the pen as an option was associated with a substantial increase in pen use, and encouragement of pen use was associated with a 100 times greater likelihood of pen use compared with respondents whose physicians did not discuss pen use or who discouraged pen use. This finding, while not unexpected, demonstrates the importance of the physician's role in patient decision-making (14).

Patients' perceptions of insulin pens were also an important predictor of pen use. The perception that ease of pen use facilitates diabetes self-care also strongly discriminated pen users from vial and syringe users. While this belief is closely related to convenience, our findings highlight the fact that it is not convenience per se that is important, but the degree to which ease of use facilitates self-care. Future research should distinguish between these closely related concepts to avoid missing a potential key advantage of insulin pens, dismissing it as a matter of mere convenience. Perception of pen cost also discriminated pen users from nonusers; in the present study, cost was the most common reason participants gave for not using a pen, with 23% of participants citing this reason.

The design of this study is cross-sectional, therefore we cannot determine the temporal ordering of the associations we report. In particular, the association between pen use and the perception that pen use facilitates diabetes self-care could reflect the fact that study participants chose to use pens because they believed these devices would facilitate diabetes regimen adherence. Alternatively, pen users may have become aware of this advantage only after they began using these devices. In either case, the implications of this association are meaningful.

The fact that physician recommendation of pen use, or even discussing insulin pens as an option, has an enormous impact on a patient's choice of insulin delivery systems has two important implications. First is the need to increase physician awareness of the potential benefits of pen use, including improved treatment satisfaction (1–6), treatment adherence (6), and clinical outcomes (2,6–7). Second, this finding suggests that physicians should talk to their patients about pen use to insure that patients are aware of this potentially beneficial insulin delivery option.

The other major implication of this study is that physicians, patients, and payors should recognize that making it easier for patients to take better care of their diabetes is not a trivial benefit. Burden of treatment is a significant barrier to improved self-care, and reducing this barrier could make an important contribution to improved diabetes outcomes.

This study was funded by an unrestricted grant to the authors from Novo Nordisk.