Development and Validation of the Chinese Insulin Medication Literacy Scale (Ch-InMLS) for Patients with Diabetes Mellitus

doi:10.21203/rs.3.rs-5299302/v1

Objective: The aim of this study was to develop a simplified insulin medication literacy scale for patients with diabetes mellitus and evaluate its psychometric properties.

Methods: We developed an initial scale based on the framework of knowledge-attitude-practice model (KAP) with skills added, and the items were developed from literature review and insulin related guidelines, semi-structured interviews, cognitive interviews. After two rounds of expert consultation and pilot survey, a final version of the scale was developed. A cross-sectional survey was conducted with 553 patients with diabetes mellitus in Zhengzhou for the psychometric evaluation. In the construct validity, the number of participants was 262 for exploratory factor analysis and 291 for confirmatory factor analysis. In the reliability, internal consistency reliability and split-half reliability were evaluated using Cronbach’s alpha coefficients.

Results: The final scale consists of 36 items with 4 dimensions, including knowledge, attitude, practice, skill. EFA suggested four factors explaining 67.556% of the total variance (Kaiser–Meyer-Olkin test = 0.944, Bartlett’s test χ2 = 7384.296, P < 0.001). The results of confirmatory factor analysis showed that the model fit the data adequately. The Cronbach’s a coefficient was 0.849 for the overall scale and ranged from 0.744 to 0.783 for domains. The Spearman-Brown split-half reliability coefficient was 0.893 for the total scale and ranged from 0.793 to 0.872 for domains. The test-retest reliability coefficient of the total scale was 0.968 and ranged from 0.880 to 0.959 for domains.

Conclusions: The scale has acceptable content validity, construct validity, and good reliability. It can be used to evaluate the level of insulin medication literacy of patients with diabetes mellitus in China.

insulin

medication literacy

diabetes mellitus

scale

reliability

validity

During the past decades, the prevalence of diabetes mellitus has increased significantly, mainly as a result of continuous rise in the incidence of type 2 DM. According to World Health Organization statistics, >422 million adults globally were suffering from DM in 2014 and a continuous rise in DM prevalence is expected^［¹^］. China is the most populous country with patients with diabetes—the prevalence of adult diabetes is 11.6% and prediabetes is 50.1% according to an epidemiological study^［²^］ . The comprehensive control rate of diabetes mellitus was 2.0%^［³^］, and only 39.7% of the estimated 113.9 million Chinese adults with diabetes have HbA1c ≤7.0%^［⁴^］, which not only brings a heavy burden, but also results in serious complications, such as diabetic nephropathy and diabetic retinopathy and increased cardiovascular mortality^［⁵^］.

Despite intensive insulin treatment can help to delay the onset of diabetes-related complications and that many patients with type 2 diabetes require insulin therapy at some stage to achieve or maintain good glycaemic control^［⁶^］, many patients are still reluctant to initiate or adherence to insulin therapy due to a variety of reasons, including the fear of addiction, injection and side effects such as hypoglycemia and weight gain, and inconvenience, and social stigma^［^7-9^］. A systematic review^［¹⁰^］ indicates that the overall prevalence of adherence to insulin therapy is remarkably low, with adherence for T1D was 52.63 %, whereas the adherence for T2D was 52.55 %.

Medication literacy is the degree to which individuals can obtain, comprehend, communicate, calculate, and process patient-specific information about their medications to make informed medication and health decisions in order to safely and effectively use their medications, regardless of the mode by which the content is delivered (e.g, written, oral, and visual)^［¹¹^］. On this basis, Pantuzza et al^［¹²^］ proposed the conceptual model for medication literacy which were consisted of four dimensions-functional literacy, communicative literacy, critical literacy, and numeracy, including the subdimensions to access, understand, evaluate, calculate, and communicate medication-related information. Compared with the concept of health literacy, medications literacy involves specific skills that are not completely covered in health literacy, such as understanding dosage instructions or information about a drug’s indication or adverse reaction. In other words, medication literacy is the specialization of health literacy in the field of medicine.

Up to now, there are a variety of generic medication literacy scales for general population, such as health literacy of pharmacy consumer questionnaire^［¹³^］, medication literacy questionnaire for discharged patients^［¹⁴^］, medication literacy in Spanish and English assessment tool^［¹⁵^］, medication health literacy measure^［¹⁶^］, recognition and addressing of limited pharmaceutical literacy interview guide^［¹⁷^］, Chinese medication literacy measure^［¹⁸^］.There are also some medication scales for particular population or disease, such as the pregnant women’s medication information literacy scale^［¹⁹^］, Chinese medication literacy scale for hypertensive patients^［²⁰^］, pharmacotherapy literacy questionnaire for patients of pre-school children^［²¹^］，medication literacy test for older adults^［²²^，²³^］. Besides, there are a few medication literacy scales for particular drug, such as medication label literacy instrument focused on non-steroidal anti-inflammatory drugs^［²⁴^］, Montana state university complementary and alternative medicine health literacy scale^［²⁵^］.

Several studies have shown that the reasons for low adherence to insulin therapy are public embarrassment^［^26,27^］, concern for hypoglycemia^［^27,28^］, negative beliefs and attitude about insulin therapy^［²⁹^］, difficulties in preparing injection, poor knowledge regarding diabetes mellitus and insulin self-injection^［³⁰^］.Therefore, systematic identification of people with limited ability to take responsibility for their insulin therapy is critical. To the best of our knowledge, most of the instruments available regarding insulin, can not cover essential aspects of insulin literacy, such as correct cognition, favorable attitude, ability to access and use information of insulin, and so on. A structured and regular assessment of patients' insulin literacy should be conducted to ensure that they are managing their insulin appropriately. This is especially relevant for type 1 diabetes who have to use insulin therapy as the mainstay, and for type 2 diabetes that cannot be controlled or tolerated by oral drugs.

Conceptual framework

This scale was developed based on Zhong’s^［²⁰^］ literacy scale for hypertensive patients and knowledge-attitude-practice model (KAP)^［³¹^］with skills added in the context of medication use. The above framework describes four main competencies for making informed medication decisions: Knowledge, Attitude, Practice and Skill.

We all know that，oral medications and insulin are effective treatments for type 2 diabetes, but for Type 2 diabetes that is not effectively controlled by oral medications as well as type 1 diabetes with absolute insulin deficiency，insulin treatment is necessary to acquire optimal glucose control and maintains HbA1c. Nonetheless, some patients have insufficient understanding of the role and status of insulin in the treatment of diabetes, for example, they believe glucose can be controlled just by taking oral medications and insulin therapy represents aggravation in disease and insulin is dependent or addictive. Therefore, understanding of related knowledge about insulin, namely knowledge, is essential for the establishment of insulin medication literacy. Besides，even with adequate knowledge，some patients interrupt or reduce the dose of insulin at will，some patients miss their insulin when they go out because of embarrassment and shame. Furthermore, there are also patients who are more likely to believe rumors about insulin on the Internet and some patients exaggerate the pain of insulin injections. Hence，the general feeling of favorableness or unfavorableness toward performing insulin management，namely attitude, is essential for the establishment of insulin medication literacy. In the international consensus study of sense of medication literacy, one of the clusters representing identified is the outcomes and goals of medication literacy^［¹¹^］. Optimal glucose control in the process of insulin use is expected to be achieved. Therefore, practices related to adherence to insulin treatment, individual management of insulin dose, communication with medical staff, as well as self-monitoring of blood glucose behavior are incorporated in the construction of insulin medication literacy. Compared to other drugs, insulin is a special dosage form that needs to be injected by the patient, the abilities and skills to use insulin effectively, safely and normatively according to printed information and education of the medical staff are one of the elements we considered to operationalize the concept of insulin medication literacy for diabetes mellitus in the present study. Specifically, skills related the technique of injection, coping with side effects, administration time and so on are incorporated in the construction of insulin medication literacy.

In this context, this study was carried out to develop a valid and reliable assessment scale of insulin medication literacy with four domains of knowledge, attitude, skill, and practice and evaluate its psychometric properties for patients with diabetes mellitus in China.

Study setting and design

This research was conducted during a cross-sectional, multi-center survey of the qualitative study. Data collection started in April 2024 to June 2024, and all participants completed informed consent.

The study was carried out in four phases. In the first phase, the Chinese Insulin Medication Literacy Scale (Ch-InMLS) for Patients with Diabetes Mellitus pool items were developed. In the second phase, content validity was conducted by twelve experts working in diabetes mellitus management or care, they were invited for a two-round Delphi. In the third phase, 200 patients with diabetes mellitus were tested in the pilot study, the items were revised or deleted by item discrimination analysis, correlation coefficient method and homogeneity test. In the fourth phase, formal investigation was conducted among 553 patients with diabetes mellitus by construct validity and reliability analysis. Figure 1 shows the process of questionnaire development.

Participants

Participants were recruited from the First Affiliated Hospital of Zhengzhou University, which is a Grade-A tertiary general hospital tertiary hospital and two community health service centers in Zhengzhou city of China. Inclusion criteria were as follows: (1) diagnosed with type 1 or type 2 diabetes according to the 1999 WHO diagnostic criteria for diabetes, (2) aged over 18 years, (3) have the ability to read and write or use WeChat, (4) have been on insulin treatment for at least four weeks, included both those who were newly diagnosed and being treated with insulin for a short period of time and those who were already on insulin treatment (once or more in a day) for a longer period of time. The exclusion criteria included: (1) patients with a history of cognitive impairment or psychiatric disease, (2) patients who declined participation, (3) patients with hearing and communication disability.

Phase 1: Development of Ch-InMLS

The development of the questionnaire consisted of the following steps: item generation, semi-structured interviews, cognitive interviews.

Item generation

For the present study, we conducted a comprehensive literature review, existing instruments not only for assessing medication /pharmacotherapy/ pharmaceutical literacy level of patients with diabetes mellitus, but also in the field of adherence/ psychological resistance/ self-management/ attitude about insulin treatment, were searched and studied in PubMed, Web of Science, Embase, Wan Fang and CNKI Data. Studies including reviews and original articles on the definition, fields and dimensions of medication literacy were also included.

The deductive method was conducted to generate items^［³²^］, which involved using current literature and insulin-related guidelines to generate items related to each domain (knowledge, attitude, practice and skill). Besides, we referred to guidelines related to insulin to enrich the connotation and composition of Ch-InMLS. Soon afterwards, an initial draft was developed in English, subsequently translated to Chinese using the back-translation method.

Semi-structured interviews

In the stage, we conducted face-to-face individual interviews on experts and patients with diabetes mellitus. Questions asked towards five endocrinologists were "What do you think are significant in insulin self-management in clinical practice?” and “What misunderstandings do you find in insulin therapy for diabetics?”. Questions asked towards eight patients were "Do you have any strengths and barriers with insulin use and what are they?". Items extracted from the review data were developed based on the four core elements of insulin literacy for diabetes mellitus patients of knowledge, attitude, skill, and practice.

In this stage, inductive method was used, combined with the above deductive method, initial drafted scale was developed.

Cognitive interviews

Cognitive interviews and pre-test of the initial drafted scale were conducted in 10 patients with diabetes mellitus through face-to-face individual interviews. The patients were invited to check the readability, comprehensibility, and response errors of the draft scale. Feedback and advice as well as questions put forward by interviewed patients on each item were recorded, and complex items with technical words which were hard to understand were replaced by more popular terms. Then, the researchers communicated with the participants and formulated an original scale based on the participants’ feedback and advice.

Phase 2: Content validity

A panel of twelve experts have been invited to appraise on the construct and items of the primary insulin literacy scale with 37 items in this study. The inclusion criteria of the expert panel: (1) with over 8 years practical experiences of diabetes mellitus management or care, (2) have intermediate or senior titles, (3) experts specializing in the development and psychometric validation of a scale, (4) are willing to participate in our study. Experts assessed expression, grammar, phrasing, item allocation of the scale according to their comprehension of the connotation of insulin medication literacy, and suggestions and the rationales were encouraged to be given. Subsequently, expurgations and revisions of items or contents have been made correspondingly.

The authority coefficient of experts were computed by their familiarity degree to the concept of insulin medication literacy and judgmental reference. An authority coefficient (Cr) of over 0.8 is considered to be acceptable^［³³^］. For experts coordination coefficient, Kendall’s coefficients of concordance (Kendall’s W) ranges from 0.40 to 0.59 indicating a grudgingly acceptable degree of chance agreement^［³⁴^］. Content validity was assessed by content validity index (CVI). Content validity index for the overall scale (S-CVI/Ave) and for each item of the scale (I-CVI) were calculated. The CVI was assessed by asking the experts to rate each item according to item’s relevance on a four-point scale, ranging from “high relevance” to “irrelevance.” Irrelevance scored one reflecting there was no correlation between the item and its belonging domain and the overall scale, and slight relevance scored two reflecting weak correlation, and relevance scored three reflecting certain correlation, and high relevance scored four indicating strong correlation. A CVI score of 0.79 or above for each item was considered to be acceptable. Items with a CVI score between 0.70 and 0.79 were revised and items with a CVI score less than 0.70 were excluded^［³⁵^］.

Phase 3: Pilot survey

To reduce the number of items, the 37-item Ch-InMLS was pre-tested in a total of 220 patients with diabetes mellitus. In this stage, a total of 200 completed questionnaires were collected back and examined validity. The response rate was 90.91%. Statistical analysis methods used for item selection were as follows.

Item discrimination analysis: total scores of collected questionnaires were ranked from high to low, among which 27% with low total scores from the lowest one were defined as low score group, and 27% with high total scores from the highest one were defined as high score group. each item in two groups was tested difference using independent t-test method and the items with no significant difference in scores between the two groups were excluded.

Correlation coefficient method: Pearson’s correlation coefficients between each item and the overall scale, and between each item and its belonging domain were calculated. Items with Pearson’s correlation coefficient r < 0.4 were considered as low correlation and were recommended to be removed.

Homogeneity test: If an item was deleted and a significant increase was present in the alpha coefficient, then deletion was considered. Besides, Communalities less than 0.2 was also considered to be removed.

Phase4: Formal investigation

To examine the psychometric properties of the tool, construct validity, criterion validity, as well as reliability were assessed.

Construct validity: Construct validity of this scale was assessed by exploratory factor analysis (EFA) which is usually used to generate the factor structure and confirmatory factor analysis (CFA) which is usually used to test the fit of the hypothetical factor structure. The EFA was conducted to extract factors by performing the principal components analysis with maximum variation method. The Kaiser–Meyer–Olkin (KMO) coefficient and Bartlett’s sphericity test were used to assess the suitability of the data. Then, the factor structure obtained from EFA was tested by CFA. Goodness-of-fit was evaluated by using the chi-square minimum/degree of freedom (χ2 /df), Root Mean Square Error of Approximation (RMSEA), Goodness of Fit Index (GFI), Comparative Fit Index (CFI), Incremental Fit Index (IFI), Parsimonious Normed Fit Index (PNFI) and Parsimonious Goodness-of-Fit Index (PGFI). The reasonable threshold levels of these indices for CFA were considered as χ 2 /df lower than 3, RMSEA，the value less than 0.08, GFI, AGFI, and IFI above 0.90, RMR less than 0.05^［³⁶^］.

The convergent and discriminant validity of the scale were assessed, and standardized factor loadings, average variance extracted (AVE), composite reliability (CR) were calculated for the final model. Convergent validity evaluates the degree of correlation of multiple items of the same domain both theoretically and practically. AVE above 0.5 and CR above 0.7 confirmed that convergent validity was satisfying. Discriminant validity indicates the level of difference between different latent variables and is valid if the average variance was greater than squared correlation coefficients^［³⁴^］.

Reliability analysis: The Cronbach’s alpha value was calculated to evaluate the internal consistency and reliability of the total scale and each dimension, and a value of greater than 0.7 was considered good internal reliability^［³⁷^］. The test–retest reliability was used to evaluate the stability of the scale and it was measured by Pearson’s correlation coefficient in 40 randomly collected patients with diabetes mellitus 2 weeks after formal investigation. A value of correlation coefficient over time more than 0.75 (P < 0.05) was considered good test-retest reliability^［³⁸^］.

Scoring Criteria for Ch-InMLS: This research scale assessed insulin medication literacy level of patients with diabetes mellitus containing four domains namely knowledge, attitude, practice and skill. For item K6 to K10，P5, S2, answering right for each item scores 2, and answering wrong or answered “I don’t know” scores 1. For item K1 to K5, A1 to A11, response option of 5-point Likert scale (totally agree, agree, not sure, disagree, totally disagree) for each item, scores of 5, 4, 3, 2, 1 were applied accordingly. For item P1 to P4, P6, P7, S1, S3 to S8, response option of 5-point Likert scale (always, often, sometimes, seldom, never) for each item, scores of 5, 4, 3, 2, 1 were also used accordingly. Besides, there were eight items in the attitude domain scoring reversely. The summed total score on this 37-item scale ranged from 37 to 164, with higher scores meaning higher insulin medication literacy level.

Development of the Ch-InMLS

A total of 40 items were screened by the literature review and insulin-related guidelines. After adding semi-structured interviews¢ results, four items were added. Subsequently, cognitive interviews were conducted among ten patients with diabetes mellitus, four items were revised. Up to now，the item pool was comprised of 44 items (see Additional file 1).

Content validity

At that stage，twelve experts were invited respectively for 2 rounds of Delphi consultations. After the first round, 5 items were deleted and 2 items were revised. After the second round, 2 items were revised (see Additional file 2). The positive coefficients of experts were 100% and 100%, indicating that experts in Delphi consultation were voluntary and active to participate to comment on the scale. The results of the expert consultation also showed that the individual authority coefficient of each expert ranged from 0.85 to 0.95, the integrated authority coefficient of all experts was 0.89, which means that the evaluation and recommendations generated from this expert panel were considerably authorized and can be trusted. Furthermore, the Kendall’s W were 0.416 (P<0.05) and 0.582(P<0.05) for two rounds, indicating a grudgingly acceptable degree of chance agreement on item appraisements.I-CVI (item-level content validity index) of each item ranged from 0.833 to 1.000, which were all above 0.79, and S-CVI (scale-level content validity index) for the total scale was 0.935, indicating a good content validity for the scale.

Pilot survey

Item discrimination analysis

Difference of each item between high score group and low score group was tested by independent t-test method (Table 1). Except for item A12, difference between high score group and low score group of the leftover 36 items were all significant (P＜0.001).

Correlation coefficient method

For knowledge domain, Pearson’s correlation coefficients between items and its belonging domain ranged from 0.747 to 0.950 (P＜0.05). For attitude domain, except for item A12, with a Pearson’s correlation coefficient of -0.064, Pearson’s correlation coefficients between other items and its belonging domain ranged from 0.734 to 0.914 (P＜0.05). For practice domain, Pearson’s correlation coefficients between each item and its belonging domain ranged from 0.695 to 0.870 (P＜0.05). For skill domain, Pearson’s correlation coefficients between each item and its belonging domain ranged from 0.701 to 0.858 (P＜0.05). Pearson’s correlation coefficients between each item and the overall scale ranged from 0.406 to 0.688 (P＜0.05), except for item A12，with a Pearson’s correlation coefficient of –0.130 (see Table 2, Table 3, Table 4, Table 5, Table 6).

Homogeneity test

Except for item A12, Cronbach' s α coefficient of the leftover 36 items after item deletion decreased (see Table 7). Besides, Except for item A5, A10, A12, Communalities of the leftover 34 items exceeded 0.2 (see Table 8).

After comprehensive consideration, item A12 was considered for deletion，the formal insulin medication literacy scale for patients with diabetes mellitus had been fulfilled, and 4 domains with 36 items were confirmed.

Formal investigation

A total of 570 patients with diabetes mellitus participated in this questionnaires，553 were collected back completely with a response rate of 97.01%. The age of the participants ranged from 25 to 84 with a mean age of 62.60 (SD=10.76). 289 out of 553 participants were male (52.3%), 151 participants had an education level of primary school or below (27.3%); 496 had been married (89.7%), 176 participants were employed (31.8%); 444 (80.3%) participants have been diagnosed as diabetes mellitus for more than 10 years; 167 (30.2%) participants had a family history of diabetes mellitus (Table 9).

Construct validity

EFA was performed to find the main factors of the tool. Construct and factor structure of this scale and of each domain were analyzed by principal component analysis with Varimax rotation. 262 collected questionnaires were randomly abstracted from 553 collected questionnaires to conduct the exploratory factor analysis for the scale.

Kaiser–Mayer–Olkin (KMO) measured for the overall scale indicated that the sample size for this scale were adequate with a value of 0.944, Bartlett’s test of sphericity was significant (χ2 = 7384.296, P < 0.001). The percentage of the total variance was 67.556%, and four domains were extracted (see Table 10). In each domain, rotation factor loadings of all items were greater than 0.6 and no item was loaded on more than one domain. The number of domains extracted by EFA and the items belonging to each domain were consistent with the initial conceptual framework. Therefore, four components of the overall scale were identified. Factor 1 was named “knowledge”, contained 10 items, factor 2 was named “attitude”, contained 11 items, factor 3 was named “practice”, contained 7 items, factor 4 was named “skill”, contained 8 items.

Based on the results of exploratory factor analysis, the rest 291 instruments of the data were used by CFA for formal comparison of models. The results of confirmatory factor analysis are shown in Table 11. Most fitting indexes of the 4 factor models reach reference values, indicating that the model has a good degree of fitting (see Table 11). The results of structure equation modeling for the CFA of scale were showed in Figure 2.

Convergent validity analysis showed that the standardized factor loading values of each item ranged from 0.647 to 0.927. The CR values ranged from 0.911 to 0.951 and the AVE values from 0.596 to 0.661. The convergent validity was acceptable. The square roots of the AVE were greater than correlations between dimensions of the scale, and it indicated reasonable discriminant validity of the scale (see Table 12).

Reliability analysis

The Cronbach’s α coefficient of the scale was calculated for internal consistency. The results showed that the Cronbach’s α coefficient of the overall scale for 36 items was 0.945, the Cronbach’s α coefficient for each dimension was 0.952, 0.947, 0.908, 0.923, indicating that internal consistency reliability of the scale was established. The total scale Spearman- Brown split-half correlation coefficient was 0.803, whilst that for each dimension was 0.925, 0.944, 0.901, 0.917, respectively. The retest reliability coefficient for the overall scale was 0.944, for each domain of the scale was 0.865, 0.845, 0.987, 0.936 (P < 0.001), suggesting the consistency of the scale and each domain over time (see Table 13).

Final scale

The scale of Ch-InMLS for patients with diabetes mellitus includes 36 items and 4 dimensions: knowledge, attitude, practice and skill (see Additional file 3), acceptable validity and good reliability and of the scale were established.

Several evaluation tool regarding insulin had been developed such as the Barriers to Insulin Treatment Questionnaire^［³⁹^］, Insulin Treatment Appraisal Scale (ITAS)^［⁴⁰^］, the Insulin Treatment Self-Management Scale^［⁴¹^］, the Chinese Attitudes to Starting Insulin Questionnaire (Ch-ASIQ)^［⁴²^］. The Barriers to Insulin Treatment Questionnaire aimed at measuring psychological barriers to insulin treatment in patients with type 2 diabetes, and all of the items focused on psychologic obstacle about insulin which was relatively one-sided when used to assess insulin medication literacy. Insulin Treatment Appraisal Scale (ITAS) was used to assess positive and negative perceptions in people with type 2 diabetes who have difficulty accepting insulin treatment. The content of ITAS placed an emphasis on subjective cognition and was oriented patients with difficulty receiving insulin therapy. The Insulin Treatment Self-Management Scale was designed to measure insulin treatment self-management of adult individuals with diabetes, and was consisted of three factors, “behavioral”, “cognitive” and “affective”. Dimension “affective” was also emotive and items such as “I am unhappy to have to use insulin”, “I delay insulin injection times with some excuses”, “I am unhappy to have to use insulin” , could not evaluate using competence of insulin objectively. The Chinese Attitudes to Starting Insulin Questionnaire (Ch-ASIQ) was designed to measure barriers to starting insulin, and was not enough to expound the connotation of insulin medication literacy.

This was the first study to report the validity and reliability of the assessment scale to comprehensively evaluate the level of insulin medication literacy for patients with diabetes mellitus in Zhengzhou. Based on the conceptual framework of knowledge-attitude-practice model (KAP) with skills added, we developed insulin medication literacy scale (Ch-InMLS) containing four dimensions, including knowledge, attitude, practice, skill. A strict, step-by-step approach, which consisted of item generation, semi-structured interviews, cognitive interviews, two rounds of expert consultation, item selection, construct validity, reliability analysis, was carried out. According to the results, the Ch-InMLS demonstrated good validity and reliability, indicating this scale can be used to assess insulin literacy literacy of patients with diabetes mellitus.

The experts consulted in semi-structured interviews had a high degree of agreement on the scale with high value of individual authority coefficient, Kendall’s W, I-CVI and S-CVI, indicating content validity was good. In the pilot study of the present study, item selection was conducted in several ways. In item discrimination analysis, item A12(“Insulin will make others perceive greater sickness”) was recommended for deletion, in correlation coefficient method item 12 was recommended for deletion once again, in homogeneity test, item A5, A10, A12 were recommended for deletion. Considering the results of statistical analysis and clinical significance, we finally deleted item A12.

In the EFA of the present study, results showed that the KMO test (KMO=0.944) and Bartlett’s test of sphericity were preferable for factor analysis. We identified four domains of knowledge, attitude, practice and skill and the four factors accounted for 67.556% of the total variance. Furthermore, CFA was conducted to verify the construct validity of the scale. Most indices meet the fitting requirements except for the GFI, AGFI were near to 0.9, RMR was near to 0.05. The χ2 /df value was less than 3. The RMSEA was less than 0.08, PGFI and PNFI were greater than 0.50. Overall, the four-domain model had acceptable construct validity, but less than perfect.

Additionally, convergent validity was identified by calculating AVE. Results demonstrated that AVE of the four domains was greater than 0.50, indicating that the amount of variance between domains and belonging items exceeded that caused by measurement errors^［⁴³^］and acceptable convergent validity was established. Discriminant validity was also assessed by calculating the average variance and squared correlation coefficient among domains. Results demonstrated that the square root of AVE was greater than all the possible two-factor correlation coefficient, illustrating that acceptable discriminant validity was established.

Cronbach’s α coefficient for the total scale and all dimensions were greater than 0.9. Spearman-Brown split-half reliability for the total scale was greater than 0.8, and Spearman-Brown split-half reliability for all dimensions were greater than 0.9, demonstrating high level of reliability and internal consistency. The test–retest reliability was tested for a time interval of two weeks The retest reliability coefficient for the overall scale and each domain of the scale were all greater than 0.7, indicating that the measurement time had little affection on the reliability of the scale, and that it had appropriate time stability. In general, our findings indicate that the scale had good reliability.

After the above four stages, the scale of Ch-InMLS for patients with diabetes mellitus includes was established including 36 items and 4 dimensions. The first named dimension was “knowledge”, which was consistent with the following statement, the first step of successful diabetes self-management is patients’ knowledge about the disease^［⁴⁴^］. The “knowledge” dimension accounted for 19.427% of the total variance, and was consisted of 10 items related to recognition of diabetes mellitus (item K1) , role of insulin in the treatment of diabetes mellitus (item K2、K3、K4、K5), basic information about the insulin being used (K6、K7、K10) , storage condition for opened and unopened insulin(K8、K9). The second named dimension was “attitude”, accounting for 20.893% of the total variance, and was consisted of 11 items related to personal beliefs about benefit of insulin (A1), psychological resistance of insulin (A2、A3), adherence of insulin (A4、A5、A6、A7、A8)，relationship between insulin and diet (A10), exercise, Internet insulin literacy (A9), pain of insulin injection (A11). Attitude served as essential mediator between knowledge and behavior, and was a significant component of the scale. The third named dimension was “practice”, accounting for 12.537% of the total variance, and was consisted of 7 items related to daily behavior under insulin treatment including blood sugar monitoring (P2), dose adjusting (P3、P7), handling of special circumstances (P1、P4、P5、P6). The fourth named dimension was “skill”, accounting for 14.699% of the total variance, and was consisted of 8 items related to injection technique (S1、S3、S5、S6、S7、S8), announcements such as injection time (S4) and treatment of hypoglycemia (S2).

Generally, the Ch-InMLS is a valid and reliable instrument to measure insulin medication literacy amongst patients with diabetes mellitus in Zhengzhou. It could be used by medical staff for the investigation on status of insulin administration. Thus, it can provide reference for simplifying insulin regimen and insulin education program, which will in turn improve insulin compliance and improve health outcomes of patients with diabetes mellitus.

Clinical trial number

Not applicable

Acknowledgements

We would like to thank the patients who participated in the study.

Author contributions

Fangying Si, Tao Feng, Xiangfen Shi, Sufang Chenand Yongjie Yang were involved in the construction of the items of the scale. Fangying Si performed the statistical analyses. All authors were involved in the discussion of the results, revision of the scale, and labelling of the subscales. Fangying Si and Tao Feng wrote the first draft of the manuscript. Sufang Chenand Yongjie Yang revised and reviewed the manuscript. All authors reviewed and approved the manuscript.

Funding

Not applicable.

Data availability

Data will be available upon request from the corresponding author.

Declarations

Ethics approval and consent to participate

Ethics was approved by the First Affiliated Hospital of Zhengzhou University Institutional Review Board approved the protocol (No. KY-2022-0470). The study was conducted following the Declaration of Helsinki. Informed consent was obtained from all participants.

Human Ethics and Consent to Participate declarations

Ethics was approved by the First Affiliated Hospital of Zhengzhou University Institutional Review Board approved the protocol (No. KY-2022-0470). The study was conducted following the Declaration of Helsinki. Informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Author details

¹Pharmacy Department, First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China

²Geriatric Endocrinology Department, First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China

Lovic, D., Piperidou, A., Zografou, I., Grassos, H., Pittaras, A., and Manolis, A. (). The Growing Epidemic of Diabetes Mellitus. Curr. Vasc. Pharmacol. 2020;18(2):104-9.
Wang, Z., Wu, Y., Wu, J., Wang, M., Wang, X., Wang, J., et al. Trends in prevalence and incidence of type 2 diabetes among adults in Beijing, China, from 2008 to 2017. Diabet Med. 2021;38(9):e14487.
Ma, A. J., Dong, J., Wei, Y. Q., Fang, K., Xie, C., Jiang, B., et al. Comprehensive control rate and related factors of diabetes mellitus in Beijing. Chin. J. Prev. Med.2020; 54(11):1283–8.
Xu, Y., Wang, L., He, J., Bi, Y., Li, M., Wang, T., et al. Prevalence and control of diabetes in Chinese adults. JAMA.2013; 310(9):948-59.
Cloete L. Diabetes mellitus: an overview of the types, symptoms, complications and management. Nurs Stand.2022; 37(1):61-6.
Riddle M. C. The current schemes of insulin therapy: Pro and contra. Diabetes. Res. Clin Pract. 2021;175, 108817.
Lee, S. H., and Yoon, K. H. A Century of Progress in Diabetes Care with Insulin: A History of Innovations and Foundation for the Future. Diabetes. Metab. J.2021; 45(5):629-40.
Dabas, H., Sarin, J., and Madhu, S. V. Insulin Adherence in Adolescents with Type 1 Diabetes Mellitus. Indian. J. Endocrinol. Metab. 2023;27(5):394-397.
Skriver, L. K. L., Nielsen, M. W., Walther, S., Nørlev, J. D., and Hangaard, S. Factors associated with adherence or nonadherence to insulin therapy among adults with type 2 diabetes mellitus: A scoping review. J. Diabetes. Complications.2023; 37(10):108596.
Boonpattharatthiti, K., Saensook, T., Neelapaijit, N., Sakunrag, I., Krass, I., and Dhippayom, T. The prevalence of adherence to insulin therapy in patients with diabetes: A systematic review and meta-analysis. Res. Social. Adm. Pharm.2024; 20(3): 255-295.
Pouliot, A., Vaillancourt, R., Stacey, D., and Suter, P. Defining and identifying concepts of medication literacy: An international perspective. Res. Social. Adm. Pharm. 2018;14(9): 797-804.
Neiva Pantuzza, L. L., Nascimento, E. D., Crepalde-Ribeiro, K., Botelho, S. F., Parreiras Martins, M. A., Camila de Souza Groia Veloso, R., et al. Medication literacy: A conceptual model. Res. Social. Adm. Pharm. 2022;18(4):2675-82.
Emmerton, L. M., Mampallil, L., Kairuz, T., McKauge, L. M., and Bush, R. A. Exploring health literacy competencies in community pharmacy. Health Expect. 2012;15(1):12-22.
Zhong, Z., Ma, G., Zheng, F., Duan, Y., Ding, S., and Luo, A. Medication Literacy in a Cohort of Chinese Patients Discharged With Essential Hypertension. Front Public Health.2020; 7:385.
Sauceda, J. A., Loya, A. M., Sias, J. J., Taylor, T., Wiebe, J. S., and Rivera, J. O. Medication literacy in Spanish and English: psychometric evaluation of a new assessment tool. J. Am. Pharm Assoc.2012; 52(6):e231-e240.
Stilley, C. S., Terhorst, L., Flynn, W. B., Fiore, R. M., and Stimer, E. D. Medication health literacy measure: development and psychometric properties. J. Nurs. Meas.2014; 22(2): 213-22.
Vervloet, M., van Dijk, L., Rademakers, J. J. D. J. M., Bouvy, M. L., De Smet, P. A. G. M., et al. Recognizing and Addressing Limited PHarmaceutical literacy: Development of the RALPH interview guide. Res. Social. Adm. Pharm. 2018;14(9): 805-811.
Yeh, Y. C., Lin, H. W., Chang, E. H., Huang, Y. M., Chen, Y. C., Wang, C. Y., et al. Development and validation of a Chinese medication literacy measure. Health Expect. 2017;20(6): 1296-1301.
Zhang, N., Wang, L., Ouyang, Y. Q., and Redding, S. Survey on medication information literacy and influencing factors among pregnant Chinese women. J. Matern. Fetal Neonatal. Med. 2021;34(10):1619-1626.
Zhong, Z., Shi, S., Duan, Y., Shen, Z., Zheng, F., Ding, S., et al. The Development and Psychometric Assessment of Chinese Medication Literacy Scale for Hypertensive Patients (C-MLSHP). Front Pharmacol. 2020;11:490.
Ubavić, S., Bogavac-Stanojević, N., Jović-Vraneš, A., and Krajnović, D. Understanding of Information about Medicines Use among Parents of Pre-School Children in Serbia: Parental Pharmacotherapy Literacy Questionnaire (PTHL-SR). Int. J. Environ. Res Public Health.2018; 15(5):977.
Pantuzza, L. L. N., do Nascimento, E., Botelho, S. F., da Rocha, A. L. P., Martins, M. A. P., do Nascimento, M. M. G., et al. Development and content validation of the medication literacy test for older adults (TELUMI). Arch. Gerontol. Geriatr.2023; 112, 105027.
Gnägi, R., Zúñiga, F., Brunkert, T., and Meyer-Massetti, C. Development of a medication literacy assessment instrument (MELIA) for older people receiving home care. J. Adv. Nurs. 2022;78(12):4210-20.
Jang, S. M., Jiang, R., Grabe, D., and Pai, A. B. Assessment of literacy and numeracy skills related to non-steroidal anti-inflammatory drug labels. SAGE. Open. Med.2019; 7:2050312119834119.
Shreffler-Grant, J., Weinert, C., and Nichols, E. Instrument to measure health literacy about complementary and alternative medicine. J. Nurs. Meas.2014;22(3):489-99.
Farsaei, S., Radfar, M., Heydari, Z., Abbasi, F., and Qorbani, M. Insulin adherence in patients with diabetes: risk factors for injection omission. Prim. Care. Diabetes. 2014;8(4):338-45.
Peyrot, M., Barnett, A. H., Meneghini, L. F., and Schumm-Draeger, P. M. Insulin adherence behaviours and barriers in the multinational Global Attitudes of Patients and Physicians in Insulin Therapy study. Diabet. Med.2012; 29(5): 682-689.
Bermeo-Cabrera, J., Almeda-Valdes, P., Riofrios-Palacios, J., Aguilar-Salinas, C. A., and Mehta, R. Insulin Adherence in Type 2 Diabetes in Mexico: Behaviors and Barriers. J Diabetes Res. 2018;3190849.
Yavuz, D. G., Ozcan, S., and Deyneli, O. Adherence to insulin treatment in insulin-naïve type 2 diabetic patients initiated on different insulin regimens. Patient Prefer Adherence.2015; 9:1225-31.
Mariye, T., Girmay, A., Birhanu, T., Tasew, H., Teklay, G., Baraki, Z., et al. Adherence to insulin therapy and associated factors among patients with diabetes mellitus in public hospitals of Central Zone of Tigray, Ethiopia, 2018: a cross-sectional study. Pan. Afr. Med J. 2019; 33, 309.
Alzghoul, B. I., and Abdullah, N. A. Pain Management Practices by Nurses: An Application of the Knowledge, Attitude and Practices (KAP) Model. Glob. J. Health Sci. 2015; 8(6): 154-60.
Elo, S., and Kyngäs, H. The qualitative content analysis process. J. Adv. Nurs. 2008; 62(1):107-115.
Chen, J., Song, Y., Ou, L., Wang, X., Wang, Y., Chen, Y., et al. Development and psychometric evaluation of a self-management behaviours scale in rheumatoid arthritis patients (RA-SMBS). BMC. Nurs.2023; 22(1):40.
Polit, D. F., Beck, C. T., and Owen, S. V. Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Res. Nurs. Health. 2007;30(4):459-67.
Narenji, M. A., Khazaee-Pool, M., and Iranpour, A. Development and Psychometric Properties of a health-promoting self-care behavior scale (HPSCB-S) in recovered patients from Drug Addiction. BMC. Public. Health. 2023; 23(1):406.
Li, J., Zhang, L., Zhou, J., Wang, G., Zhang, R., Liu, J., et al. Development and validation of self-management scale for tuberculosis patients. BMC. Infect. Dis. 2022; 22(1):502.
Khazaee-Pool, M., Moridi, M., Ponnet, K., Turner, N., and Pashaei, T. Psychometric properties of the Persian version of the Time to Relapse Questionnaire (TRQ) in substance use disorder. Am J Drug Alcohol Abuse.2016; 42(6):682-8.
Koo, T. K., and Li, M. Y. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J. Chiropr. Med.2016; 15(2):155-63.
Petrak, F., Stridde, E., Leverkus, F., Crispin, A. A., Forst, T., and Pfützner, A. Development and validation of a new measure to evaluate psychological resistance to insulin treatment. Diabetes Care. 2007;30(9):2199-204.
Snoek, F. J., Skovlund, S. E., and Pouwer, F. Development and validation of the insulin treatment appraisal scale (ITAS) in patients with type 2 diabetes. Health Qual Life Outcomes. 2007; 5:69.
Karahan Okuroglu, G., Karaçanta Atbaş, S., and Ecevit Alpar, Ş.Development, reliability, and validity of the Insulin Treatment Self-Management Scale. Int. J. Nurs. Pract. 2020; 26(5):e12814.
Fu, S. N., Chin, W. Y., Wong, C. K., Yeung, V. T., Yiu, M. P., Tsui, H. Y., et al. Development and validation of the Chinese Attitudes to Starting Insulin Questionnaire (Ch-ASIQ) for primary care patients with type 2 diabetes. PLoS. One.2013; 8(11): e78933.
Acosta-Prado, J. C., López-Montoya, O. H., Sanchís-Pedregosa, C., & Zárate-Torres, R. A. Human Resource Management and Innovative Performance in Non-profit Hospitals: The Mediating Effect of Organizational Culture. Front Psychol.2020; 11:1422.
Jiang, T., Li, A., Zhang, M., Zhou, Z., Wang, L., Zhang, X., et al. Measuring Self-management Among People with Diabetes Mellitus: A Systematic Review of Patient-Reported Diabetes-Specific Instruments in English and Chinese. Adv Ther.2023; 40(3), 769-813.

TABLE 1 Independent t-test between high score group and low score group

Item	t	sig(bilateral)
K1	12.219	0.000
K2	13.279	0.000
K3	10.139	0.000
K4	10.533	0.000
K5	13.849	0.000
K6	10.894	0.000
K7	9.808	0.000
K8	8.831	0.000
K9	12.761	0.000
K10	10.872	0.000
A1	7.895	0.000
A2	11.895	0.000
A3	9.297	0.000
A4	8.899	0.000
A5	6.037	0.000
A6	9.174	0.000
A7	9.846	0.000
A8	8.664	0.000
A9	9.318	0.000
A10	4.778	0.000
A11	10.387	0.000
A12	-1.609	0.110
P1	6.836	0.000
P2	9.850	0.000
P3	9.974	0.000
P4	11.898	0.000
P5	13.654	0.000
P6	12.185	0.000
P7	10.403	0.000
S1	10.391	0.000
S2	13.654	0.000
S3	13.727	0.000
S4	8.402	0.000
S5	9.159	0.000
S6	5.899	0.000
S7	12.045	0.000
S8	10.943	0.000

TABLE 2 Pearson’s correlation coefficients between item K1 to K10 and total score of Knowledge domain

	P	sig(bilateral)
K1	0.872	0.000
K2	0.950	0.000
K3	0.842	0.000
K4	0.848	0.000
K5	0.941	0.000
K6	0.786	0.000
K7	0.797	0.000
K8	0.747	0.000
K9	0.803	0.000
K10	0.806	0.000

TABLE 3 Pearson’s correlation coefficients between item A1 to A12 and total score of Attitude domain

	P	sig(bilateral)
A1	0.765	0.000
A2	0.914	0.000
A3	0.846	0.000
A4	0.828	0.000
A5	0.734	0.000
A6	0.823	0.000
A7	0.748	0.000
A8	0.821	0.000
A9	0.835	0.000
A10	0.738	0.000
A11	0.790	0.000
A12	-0.064	0.000

TABLE 4 Pearson’s correlation coefficients between item P1 to P7 and total score of Practice domain

	P	sig(bilateral)
P1	0.695	0.000
P2	0.765	0.000
P3	0.790	0.000
P4	0.803	0.000
P5	0.870	0.000
P6	0.816	0.000
P7	0.811	0.000

TABLE 5 Pearson’s correlation coefficients between itemS1 to S8 and total score of Skill domain

	P	sig(bilateral)
S1	0.813	0.000
S2	0.858	0.000
S3	0.826	0.000
S4	0.797	0.000
S5	0.746	0.000
S6	0.701	0.000
S7	0.845	0.000
S8	0.797	0.000

TABLE 6 Pearson’s correlation coefficients between each item and total score of the scale

	P	sig(bilateral)
K1	0.666	0.000
K2	0.688	0.000
K3	0.586	0.000
K4	0.602	0.000
K5	0.670	0.000
K6	0.569	0.000
K7	0.528	0.000
K8	0.536	0.000
K9	0.608	0.000
K10	0.566	0.000
A1	0.559	0.000
A2	0.672	0.000
A3	0.627	0.000
A4	0.576	0.000
A5	0.437	0.000
A6	0.615	0.000
A7	0.580	0.000
A8	0.595	0.000
A9	0.612	0.000
A10	0.406	0.000
A11	0.668	0.000
A12	0.130	0.000
P1	0.470	0.000
P2	0.554	0.000
P3	0.565	0.000
P4	0.653	0.000
P5	0.643	0.000
P6	0.605	0.000
P7	0.601	0.000
S1	0.523	0.000
S2	0.582	0.000
S3	0.641	0.000
S4	0.484	0.000
S5	0.546	0.000
S6	0.450	0.000
S7	0.541	0.000
S8	0.518	0.000

TABLE 7 Cronbach' s α coefficient after item deletion for homogeneity test

	Cronbach' s α for the scale	Cronbach' s α coefficient after item deletion
	0.928
K1		0.924
K2		0.924
K3		0.925
K4		0.925
K5		0.924
K6		0.926
K7		0.927
K8		0.927
K9		0.926
K10		0.927
A1		0.926
A2		0.924
A3		0.925
A4		0.925
A5		0.927
A6		0.925
A7		0.925
A8		0.925
A9		0.925
A10		0.928
A11		0.924
A12		0.935
P1		0.927
P2		0.926
P3		0.926
P4		0.924
P5		0.926
P6		0.925
P7		0.925
S1		0.926
S2		0.926
S3		0.925
S4		0.926
S5		0.926
S6		0.927
S7		0.926
S8		0.926

TABLE 8 Communalities of each item

	Extract
K1	0.562
K2	0.611
K3	0.450
K4	0.467
K5	0.608
K6	0.465
K7	0.408
K8	0.408
K9	0.506
K10	0.439
A1	0.222
A2	0.315
A3	0.281
A4	0.225
A5	0.110
A6	0.268
A7	0.248
A8	0.250
A9	0.259
A10	0.087
A11	0.342
A12	0.042
P1	0.216
P2	0.286
P3	0.320
P4	0.423
P5	0.436
P6	0.364
P7	0.332
S1	0.301
S2	0.385
S3	0.450
S4	0.251
S5	0.304
S6	0.210
S7	0.350
S8	0.299

TABLE 9 Basic demographic characteristic

Items	Group	N	％
Age(years)*	18~45	23	4.2
	46~60	179	32.4
	＞61	351	63.5
Gender	Male	289	52.3
	Female	264	47.7
Education level	Primary and below	151	27.3
	Junior middle school	197	35.6
	High school	136	24.6
	College degree and above	69	12.5
Marital status	Unmarried	22	4.0
	Married	496	89.7
	Divorced/Widowed	35	6.3
Occupational status	Employed	176	31.8
	Unemployed/Retired	377	68.2
Duration of diabetes	≤3 years	23	4.2
	4~9 years	86	15.6
	≥10 years	444	80.3
Registered residence	Urban	325	58.8
	Countryside	228	41.2
Annual household income Chinese CNY (¥)	<10,000/year	43	7.8
	10,000~29,999/year	235	42.5
	30,000~49,999/year	154	27.8
	50,000~99,999/year	121	21.9
	≧100,000/year
Family history of diabetes mellitus	Yes	167	30.2
	No	386	69.8

*The mean for age was 62.60 years with a standard deviation of 10.76.

Table 10 Exploratory factor analysis on the scale of Ch-InMLS for patients with diabetes mellitus (n=262)

Item	Factors
Item	1	2	3	4
K5	0.888
K2	0.874
K7	0.789
K4	0.786
K1	0.786
K3	0.782
K9	0.777
K6	0.773
K10	0.767
K8	0.726
A2		0.905
A3		0.838
A4		0.833
A9		0.825
A8		0.824
A6		0.818
A5		0.787
A1		0.781
A11		0.780
A10		0.770
A7		0.724
P5			0.817
P7			0.769
P6			0.765
P3			0.756
P2			0.726
P4			0.704
P1			0.688
S2				0.856
S1				0.788
S7				0.784
S3				0.766
S8				0.754
S4				0.752
S5				0.746
S6				0.664
Eigenvalues	6.216	12.359	2.638	3.108
Explained Variations (%)	19.427	20.893	12.537	14.699
Total Explained Variations (%)	67.556

K, knowledge domain; A, attitude domain; S, skill domain; P, practice domain; KMO measure of sampling adequacy value=0.944, Bartlett’s test: c2 (chi square test value) =7384.296; df (degree of freedom) =630; P=0.000.

Table 11 The results of fitting indices of confirmatory factor analysis of four-domain model of Ch-InMLS for diabetes mellitus (n=291)

Parameters	Four-domain model	Reference value
c2/df	1.237	≤3
GFI	0.878	＞0.90
AGFI	0.862	＞0.90
RMR	0.056	＜0.05
IFI	0.982	＞0.90
RMSEA	0.029	＜0.08
PGFI	0.916	＞0.50
PNFI	0.850	＞0.50

Table 12 Convergent and discriminant validity of the scale

Total scale/ Dimensions	Sted	Square correlation	CR	AVE	square root of AVE
K1	0.862	0.743	0.951	0.661	0.813
K2	0.910	0.828
K3	0.796	0.634
K4	0.783	0.613
K5	0.870	0.757
K6	0.812	0.659
K7	0.767	0.588
K8	0.753	0.567
K9	0.792	0.627
K10	0.770	0.593
A1	0.825	0.681	0.944	0.606	0.778
A2	0.893	0.797
A3	0.751	0.564
A4	0.817	0.667
A5	0.647	0.419
A6	0.795	0.632
A7	0.666	0.444
A8	0.790	0.624
A9	0.769	0.591
A10	0.754	0.569
A11	0.820	0.672
P1	0.663	0.440	0.911	0.596	0.772
P2	0.759	0.576
P3	0.726	0.527
P4	0.769	0.591
P5	0.927	0.859
P6	0.793	0.629
P7	0.740	0.548
S1	0.770	0.593	0.928	0.616	0.785
S2	0.859	0.738
S3	0.767	0.588
S4	0.678	0.460
S5	0.815	0.664
S6	0.809	0.654
S7	0.822	0.676
S8	0.746	0.557

Table 13 The reliability coefficients for the total scale and each dimension of Ch-InMLS scale for diabetes mellitus (n=553)

Domains	Items	Cronbach’s α coefficient	Spearman-Brown split-half reliability	Test-retest reliability
K	10	0.952	0.925	0.865
A	11	0.947	0.944	0.845
P	7	0.908	0.901	0.987
S	8	0.923	0.917	0.936
Ch-InMLS	36	0.945	0.803	0.944

No competing interests reported.

Development and Validation of the Chinese Insulin Medication Literacy Scale (Ch-InMLS) for Patients with Diabetes Mellitus

Status:

Version 1

Abstract

Figures

INTRODUCTION

METHODS

RESULTS

DISCUSSION

CONCLUSION

Declarations

References

Tables

Additional Declarations

Status:

Version 1