Scale Development

Purpose of the HFS and Theoretical Background

The Hypoglycemia Fear Survey (HFS) was originally designed to measure the degree of fear or anxiety associated with episodes of hypoglycemia and their negative consequences in adults with type 1 diabetes. The survey was first published in 1987.⁴ The HFS is based on cognitive-behavioral theories of anxiety that include an anxious affective and an avoidant behavioral component. It is important to note, that some level of anxiety about hypoglycemia is, not only normal, but also adaptive. People with diabetes who take insulin and other blood glucose-lowering medications need to be aware and stay vigilant for signs and risk factors for hypoglycemia, and perform appropriate treatment behaviors to avoid severe episodes. For some individuals, having too little fear of hypoglycemia can be a problem and increase the risk of severe episodes, while other people may experience very high levels of anxiety that can have a negative effect on diabetes outcomes and quality of life.

Scale design and development

Subscales

All versions of the HFS are comprised of two subscales: The Behavior Subscale (HFS-B) and the Worry Subscale (HFS-W). In the current adult version of the HFS, the HFS-II, the first 15 items assess behaviors that individuals may engage in to avoid hypoglycemia and its potential negative consequences (HFS-B). Three of these items reflect diabetes management behaviors individuals may perform to reduce the chance of hypoglycemia (e.g. reduce insulin when BG is low). Three other items address behaviors aimed at keeping BG levels in a higher range in order to avoid hypoglycemia (e.g. keeping BG higher when in social situations). The remaining nine items reflect behaviors aimed at avoiding high-risk situations (e.g. limiting out of town travel). The last 18 items of the HFS-II assess specific worries related to various aspects of hypoglycemia (HFS-W). These include worries about episodes occurring while an individual is alone or when there is no food or drink available for treatment, worries about having an accident or physical injury, and worries about being unable to function normally

Development of the Scale

The procedure for the original development of the HFS began with interviews of diabetes health care providers about their impressions of patients’ concerns over hypoglycemia. Then twenty individuals with type 1 diabetes were interviewed for possible items using an open-ended question format (e.g., “What bothers you about hypoglycemia?” and “What do you do to avoid or cope with hypoglycemia?”). From this qualitative data, the authors developed 34 items to be rated on a Likert scale from 1 (“never”) to 5 (“always”). These items represented two clusters intended to measure patients’ behavioral avoidance and affective FoH. After administering the preliminary scale to a group of 12 patients and diabetes health-care providers, redundant items (inter-item correlation >.80) were deleted or revised, resulting in 29 items.

This revised version was again clustered into two subscales: behavior and worry, and administered to 35 individuals with type 1 diabetes. After analyses of responses from this second administration, seven items were rewritten to clarify the concepts and two items were dropped altogether. The final HFS was comprised of 10 behavior items (items 1-10) and 17 worry items. Finally, to assess generalizability, the HFS was administered to 158 individuals with type 1 diabetes from four different geographic areas across the United States. Internal reliability was good: Cronbach’s alpha for the entire scale was .90 - .60 for the Behavior Subscale, and .89 for the Worry Subscale, and all items were positively correlated with the Total Scale.

Subsequent revisions of the HFS

In subsequent revisions of the HFS, more items were deleted from the original scale due to low response variance and other items were reworded to clarify their meaning. The final instrument after this revision, the HFS-I, contained a total of 23 items: 10 Behavior items and 13 Worry items. Over the next several years, the HFS-I was used in dozens of studies investigating topics that included: differences in FoH between people with type 1 and type 2 diabetes , factors that increase or decrease FoH, and the relationship between FoH and HbA1c levels .

The HFS was revised again in 1998 stemming from feedback from people with diabetes and colleagues in diabetes research, bringing the total to 33 items: 15 Behavior items and 18 Worry items. The Likert rating scale was also revised from a scale of 1-5 to 0-4 where 0 represents “Never” and 4 represents “Almost Always.” This revised scale, the HFS-II, has continued to be used since these changes. Since its development, the HFS-II has been used in numerous scientific studies and has been demonstrated to be sensitive to change in a broad range of clinical trials including educational and behavioral interventions, diabetes medications and technologies, and beta cell transplant procedures. It has also been used in studies of people with type 2 diabetes as well as type 1. At present, there are over 70 linguistically-validated translations of the adult version of the HFS-II. Click here to see a list of available translations.

Short Form of the HFS-II (Adult)

The HFS-II (Adult) was adapted into a shorter 11-item scale in order to decrease patient/participant burden and enhance the instrument’s utility in clinical and research settings. The short form is composed of five Behavior subscale items and six Worry subscale items, and was developed using item response theory methods from a sample of 487 type 1 and type 2 participants.⁸ Subsequently, the short form was validated on a sample of 2,718 type 1 and type 2 participants taking part in a large observational study of the frequency of hypoglycemia in France. Exploratory factor analysis demonstrated that the short form followed the 3-factor structure of the long form. Overall, the short form was highly correlated with the long form of the scale on both the total scale and each subscale (Pearson’s R > .89), and maintained high internal consistency on each subscale and the total scale (> .76).

Instructions for Use of the HFS-II

Altering the time frame for the HFS-II

The HFS-II instructions provide a timeframe for the recall of hypoglycemia-related thoughts, feelings and behaviors (e.g. during the last 6 months) This timeframe can be adjusted to fit the clinical or research needs of the user. For example, for a study investigating a 4-week intervention, the instructions can be changed to “during the last 4 weeks. The survey can also be administered via paper and pencil, or electronically.

Use of Part II of the HFS-II

Part II of the HFS is a questionnaire regarding the person’s recent experiences with hypoglycemic episodes. This part of the survey is provided if studies (or clinics) are not going to have access to information about both the frequency of hypoglycemia and the degree of emotional upset associated with past episodes. It is impossible to interpret the results of the HFS-II with no information about an individual’s exposure to and experience with hypoglycemia. For example, an individual who has experienced recent severe episodes of hypoglycemia associated with serious consequences such as physical injury will most likely show relatively high scores on the HFS-II. On the other hand, if an individual has very few episodes of hypoglycemia, including mild episodes, high scores would not be expected and, if seen, may be indicative of some sort of problem with diabetes-related anxiety.

Scoring the HFS-II

Scoring manuals are available for both the adult and pediatric versions of the HFS.