Development of a new pregnancy informational and decisional needs survey for women with physical disabilities

Abstract

Background:

While the literature on pregnancy and disability is growing, generating important knowledge of barriers to care and health risks, there is limited literature on pregnancy decision-making and informational needs. Such knowledge is critical for the development of interventions to mitigate the challenges women with disabilities and health care providers face in making this important decision.

Objective: /Hypothesis:

Develop a survey of pregnancy decisional and informational needs of women with physical disabilities.

Methods:

We developed a conceptual framework that guided item writing, defined the characteristics of the target population, engaged stakeholders with disabilities to collaborate with investigators, and tested the interpretability and relevance of items, and preferred mode of responding (online vs. telephone). A total of 123 women participated in cognitive interviews (N = 13), mode testing (N = 10), or survey data collection (N = 114).

Results:

Instructions, three screening questions, and 156 items were tested in one round of 17 cognitive interviews; 25 items were deleted, 94 were revised, and 37 items were retained without changes. The final version of the survey included six sub-sections: the experience of making a decision; information about pregnancy and disability; things affecting a decision; knowing what is important; support for making a decision; and working with health care providers.

Conclusions:

The overall performance of the survey supported its content validity and utility as a useful way to gather information about the pregnancy decision-making experience and informational needs of women with physical disabilities.

Introduction

A decision to pursue a pregnancy for any woman should involve well-informed deliberation with her health care providers. The need for high quality decision-making is even more important for women with physical disabilities, characterized by mobility impairment, given a greater the potential risks and tradeoffs in health and birth outcomes (e.g., pre-term birth,¹ low birth weight,^2,3 miscarriage,^4,5 hospitalizations during pregnancy,¹ and prolonged postpartum stays⁶) they may encounter. In addition, the receipt of family planning services are highly variable and women with physical disabilities with low education, low income and who are unemployed are particularly disadvantaged.⁷ Health care provider limited expertise in disability^8,9 can often lead to inaccurate or inappropriate guidance on risks and potential complications.¹⁰ A lack of clinical guidelines,¹¹ compounded by stigma around pregnancy and disability,^12,13 contribute to a compromised decision-making process for many women with physical disabilities and their health care providers.

While the literature on pregnancy and disability is growing, generating important knowledge of barriers to care and health risks, there is limited literature on pregnancy decision-making and informational needs in this population of women. Such knowledge is critical for the development and implementation of interventions to mitigate the numerous challenges women with disabilities and their health care providers face in making this important decision. The purpose of this paper is to describe the development of a survey designed specifically for women with physical disabilities to assess their pregnancy decision-making and informational needs to help generate this knowledge. Subsequent publications will describe the results of this new survey and the development and pilot testing of the new pregnancy decision-making tool.

Methods

To help ensure the content validity^14,15 of the new survey, we developed a conceptual framework, defined the characteristics of the target population, engaged women stakeholders with disabilities (AW, MB, BO) to collaborate with investigators (CZK, SE, HH) throughout the development process, and tested the interpretability of items and their relevance to pregnancy decision-making of women with physical disabilities.

Constructing the conceptual framework around pregnancy and disability

We used the Ottawa Framework for Decision Support¹⁶ as the theoretical foundation of this new survey. This model targets determinants of decisions that may be potentially modifiable by a decision-making tool, but are currently suboptimal for patients and health care providers due to factors such as inadequate knowledge, high uncertainty, or biased perceptions of others. Jacobsen and O’Connor’s work on developing decisional needs assessment¹⁷ provides several adaptable templates for assessing specific decision-making needs. Using this framework as a conceptual anchor and the templates as a starting point, we organized the survey content around five domains. Decisional conflict refers to uncertainty about a course of action to take when making a choice that involves risk, loss, regret, or challenge to personal life values. Knowledge and expectations refers to the cognizance of the situation, options, and outcomes (knowledge) and the perceived likelihood or probability of outcomes of options (expectations). Values refers to the personal importance of the outcomes of options. Support and resources includes others’ opinions, pressures, experiences, motivations, and external support.

In the Ottawa Framework for Decision Support, these constructs are framed around health decisions and, most often, the treatment of a disease or health condition where two or more options often exist. In contrast, pregnancy is not a health decision per se, although it can have significant health implications, nor is it a decision driven by treatment of disease or health condition. Nevertheless, with some slight modifications, these constructs were informative for constructing the conceptual framework to guide the survey’s content development. Personal and clinical characteristics focuses on those factors most relevant to the decision. Based on other work to develop a conceptual framework of reproductive health in women with physical disabilities¹⁸ and feedback from stakeholders, we also included a section on working with health care providers and what advice respondents would give other women with disabilities and health care providers about making this decision.

Defining the Survey’s purpose and target population

Initially, we defined the target population as women who had been pregnant, decided not to get pregnant, and who were currently deciding in the target population. However, given many pregnancies are unplanned or miscarried, we also included women who had an unplanned pregnancy and decided to carry to term or who had a miscarriage of a planned pregnancy. Women who had never considered a pregnancy, had never been pregnant or who may consider a pregnancy in the distant future, but had not yet spent time contemplating the decision were not considered as part of the target population. Defining the time frame of considering a future pregnancy most relevant for the survey was particularly challenging. We considered referring to a specific time frame (e.g., 6 months), but decided that was likely to be too restrictive; “near future” was more open but still challenging to define. To help determine the best way to present a time frame in the survey, we included this question in item interpretability testing. No time limit was set on when a past pregnancy had occurred.

Item development

For guidance on developing items and writing the initial drafts, we used Jacobsen and O’Connor’s templates for assessing specific decision-making needs.¹⁹ These templates include both closed- and open-ended items which enabled us to gather specific information as well as more narrative information. Items and instructions were developed in iterative rounds involving all investigators and steering committee members. While many items from the template were a good starting point and easily modifiable, a number were phrased in ways that were either not applicable or awkward for a decision about pregnancy. These were most often questions related to treatment decisions. For example, referring to “pros and cons” or “advantages and disadvantages” of a decision, selecting “an option”, or “wavering between choices” did not translate well for pregnancy. In contrast, items about factors making the decision difficult such as lacking information or feeling unsupported could be easily framed for pregnancy with very little or no modification.

Each section included one or more open-ended items. For all but a handful of closed-ended items, 5-point Likert scales of intensity (not at all to very much) were used as the response set. Because the survey was designed for women with physical disabilities and not a specific condition or injury, an option of “not applicable” was added to the response sets of most closed-ended items that we expected might not be relevant to all respondents (e.g., references to having a partner, passing on a genetic disorder). We also included at least one open-ended item for additional information that was important but not included in the preceding items, which helped to identify gaps in item coverage. Three additional questions were written to identify the pregnancy decision-making status of respondents: considering a future pregnancy, had been pregnant before (planned or unplanned) or had made a decision to never get pregnant. The item development process took place from June to October 2018. The survey’s sub-sections, alignment with the Ottawa Framework for Decision Support, and how the topic was described for the respondents are given in Table 1.

Table 1.

Initial survey sub-sections, alignment with ottawa decision support framework, and summary of feedback.

Sub-Section	Ottawa Framework for Decision Support Alignment	Description for Respondent
Experience of making a decision about pregnancy	Decisional conflict	This first set of questions are about your overall experience of making a decision about pregnancy.
Information about pregnancy and disability	Knowledge and expectations	These next questions are about information that you need/needed to help inform your decision about pregnancy. First, you will check anything on the list that applies/applied to you and add anything not on the list by checking one or more of the “other” items at the end. Next, for each item you checked and ones you added, you will rate its importance, helpfulness, ease or difficulty getting, and where you got it.
Things affecting a decision about pregnancy	Decisional conflict	These next questions are about things that can make a decision about pregnancy harder or easier or not affect it at all. For each question, “your decision making” refers to your decision specifically about pregnancy.
Knowing what is important	Values	These next questions are about the importance of certain things when making a decision about a pregnancy.
Support for making a decision about pregnancy	Support and resources	These next questions are about support and resources when making a decision about pregnancy. “Support from others” refers to how much people in your life, like a partner, family, friends and health care providers, give/gave you emotional support and practical support. “Resources” include things like information and advice from various sources.
Communicating with health care providers about pregnancy	Not applicable	These next questions are about your experience talking with health care providers to get information, advice, or any other guidance for making a decision about getting pregnant.

Testing interpretability and relevance

Once items and instructions were in final draft form, they were submitted to cognitive testing, a standard technique in the development of surveys that explicitly focuses on cognitive processes that respondents use to answer.²⁰ Feedback from these interviews was used to help ensure items were interpreted as intended and relevant to women with physical disabilities. Interviews were organized by sub-sections and tested with a minimum of 5 respondents²¹ during a telephone interview. Three main topics were addressed for each item: 1) what they were thinking about when answering; 2) ease of answering the question and sufficiency of the response set; and 3) importance of the question for women with disabilities. For selected items, questions about potentially problematic words or phrases identified in item writing were also addressed. All interviews were digitally audio-recorded for back up documentation. One or more of eight common problems based on other cognitive interviewing studies^22–24 were applied to items based on cognitive interview feedback (see Supplemental Materials.) Interviews took place between October and November 2018.

Online survey form and survey mode testing

The online form of the survey was administered using REDCap,²⁵ an online data capture and management system. We used text piping that customized each item (most often the item’s stem) based on the time frame appropriate for the respondent determined during screening. Mode testing refers to checking whether the method of data collection or survey administration (e.g., telephone, face to face, paper, mobile apps, or online) has differential effects on responding. We recruited the first 10 respondents to complete telephone interview and online versions, randomizing their order, of the survey forms to test for possible mode effects. Once mode testing was completed, we compared responses of the two modes and conducted a brief follow up interview to understand possible reasons for differences in responses.

Approach to quantitative analysis

The purpose of this survey is to explore the decision-making experience of respondents rather than produce a score or quantify a construct or set of constructs. As such, the lack of a single summary score or sub-scores, use of closed and open-ended items, a “not applicable” response option, and skip logic made it difficult to evaluate common psychometric characteristics, such as factor structure. Rather, we evaluated the performance of items using a mix of quantitative and qualitative approaches. Quantitatively we examined the distributions of responses, the frequency of missing and “not applicable” responses, and calculated Cronbach’s alpha²⁶ for sets of closed-ended items with values > 0.70 as preliminary evidence of internal consistency. To further evaluate content validity and item coverage, we qualitatively examined open-ended responses at the end of each section that referred to important and/or consistently mentioned concepts not included in the preceding closed-ended items.

Results

Sample characteristics

A total of 123 women participated in cognitive interviews (N = 13), mode testing (N = 10), or the survey (N = 114; see Table 2). Only two of the 114 participants did not complete their survey form (completing 20% and 50%, respectively; less than 2% of the sample). Women had to have a mild, moderate or severe physical disability defined as a loss of physical function or mobility that restricts the individual from one or more important life activities assessed with items from the Behavioral Risk Factors Surveillance System.^27,28 We used three items that assess whether a person is limited in any way because of a mobility impairment and, if so, their need for assistance with daily life activities and/or personal care. Among the most commonly reported disabling conditions were musculoskeletal disorders (n = 19), spina bifida (n = 20), cerebral palsy (n = 19), Ehlers-Danlos syndrome (n = 11), Arthrogryposis multiplex congenital (n = 11), and Multiple Sclerosis (n = 10). Less common causes of disability included autoimmune disorders (e.g., lupus, rheumatoid arthritis), muscular dystrophy, neurological injuries or disorders, limb loss, and fibromyalgia. Nearly 60% of the sample (n = 74) was born with a disability. The largely non-Hispanic Caucasian sample was relatively well educated with the majority completing at least some post-secondary education. This study was approved for an IRB exemption #2 by the University of XXXX IRBMED; women were provided with all elements of consent and verbally agreed to participate in cognitive interviews and survey data collection (see Table 3).

Table 2.

Cognitive interview^a, mode effect^b and survey^a participant characteristics.

	Cognitive Interviews (N = 13)	Mode Effects (N = 10, subset of Survey N = 114)	Survey (N = 114)
Age at disability onset (if not birth; M, SD; Range)	23.3 (9.6); 1–32	20.6 (7.0); 13–27	26.0 (9.2); 9–46
Current age (M, SD; Range)	33.0 (7.3); 20–45	35.2 (10.9); 26–62	35.6 (9.5); 20–75
Severity of disability
Mild	2 (15.4%)	2 (20.0%)	21 (18.4%)
Moderate	4 (30.8%)	4 (40.0%)	52 (45.6%)
Severe	7 (53.8%)	4 (40.0%)	41 (36.0%)
Pregnancy Decision-making Statusb
Was pregnant in the past while disabled		3 (30.0%)	38 (33%)
In the process of making a decision about pregnancy (never been pregnant while disabled)		4 (40.0%)	31 (27.8%)
Made a decision to never get pregnant while disabled		2 (20.0%)	20 (17.4%)
In the process of making a decision about pregnancy (has been pregnant while disabled)		1 (10.0%)	12 (10.4%)
Currently pregnant (has been pregnant before while disabled)		0 (0%)	5 (4.3%)
Made a decision to never get pregnant again while disabled		0 (0%)	5 (4.3%)
Actively trying to get pregnant		0 (0%)	2 (1.8%)
Currently pregnant (never been pregnant while disabled)		0 (0%)	1 (0.9%)
Highest education completed
High school/GED	2 (15.4%)	0 (0%)	10 (8.8%)
Associates degree	0 (0%)	1 (10.0%)	9 (7.9%)
Some college (no degree)/Other	0 (0%)	0 (0%)	31 (27.1%)
Bachelor’s degree	6 (46.2%)	7 (70.0%)	33 (28.9%)
Postgraduate degree	5 (38.5%)	2 (20.0%)	31 (27.2%)
Marital statusb
Single, never married	4 (30.8%)	1 (10.0%)	20 (17.5%)
Married	4 (30.8%)	3 (30.0%)	55 (48.2%)
Significant Other	4 (30.8%)	4 (40.0%)	29 (25.4%)
Divorced/Separated	1 (7.7%)	2 (20.0%)	10 (8.8%)
Living situation
Live alone	1 (7.7%)	1 (10.0%)	78 (68.0%)
Live with spouse/significant other	8 (61.5%)	7 (70.0%)	16 (14.0%)
Live with parents	3 (23.1%)	0 (0%)	8 (7.0%)
Live with children	1 (7.7%)	1 (10%)	8 (7.0%)
Other	0 (0%)	1 (10%)	4 (3.5%)
Race and Ethnicity
White or Caucasian	12 (92.3%)	9 (90%)	94 (82.5%)
African American	0 (0%)	1 (10%)	13 (11.4%)
Asian	1 (7.7%)	0 (0%)	1 (0.9%)
American Indian/Alaska Native	0 (0%)	0 (0%)	1 (0.9%)
Native Hawaiian/Other Pacific
Islander	0 (0%)	0 (0%)	1 (0.9%)
Hispanic Ethnicity	2 (15.4%)	1 (10.0%)	8 (7.0%)
Living in a medically underserved area	1 (7.7%)	2 (20.0%)	13 (11.4%)
Living in a health professional shortage area Primary care	8 (61.5%)	3 (30.0%)	57 (50.0%)
Dental care	6 (46.2%)	4 (40.0%)	38 (33.3%)
Mental health care	8 (61.5%)	4 (40.0%)	50 (43.9%)

^a4 women completed both cognitive interview and survey and are represented in those respective columns; 2 completed all three study activities and are represented in all columns.

^bThese data were not collected from cognitive interview participants.

Table 3.

Open ended comments regarding missing concepts/items (N = 114).

Sub-Section	No comment given (N, %)	“No” response (N, %)	Some comment (N, %)
Experience of making a decision about pregnancy	34 (30%)	15 (13%)	65 (57%)
Things affecting the decision about pregnancy	49 (43%)	25 (22%)	40 (35%)
Things that are important to a decision about pregnancy	62 (54%)	27 (24%)	25 (22%)
Working with health care providers about pregnancy	50 (44%)	23 (20%)	41 (36%)

Cognitive Interview Feedback

The instructions, three screener questions, and 156 items were tested in one round of 17 cognitive interviews (see Fig. 1). After review of feedback, 25 items were deleted, 94 were revised, and the remaining 37 items were retained without changes. The most common problems were being unclear in their intent or context (n = 26), wording or tone of the item (n = 11), an inadequate response set (n = 9), or redundancy (n = 7). We specifically sought feedback on the screening question, “Are you thinking about getting pregnant at some point in the near future?” as to whether a specific timeline should be defined for “near future” or if “actively planning” was preferable. The majority of women felt that it was not necessary to define a time interval and that actively planning was unnecessarily restrictive. Instructions and the other two screener questions were felt to be clear and were not changed. Items were nearly unanimously felt to be important to pregnancy decision-making of women with physical disabilities; in a handful of cases, when an item was not felt to be important it was usually because the respondent did not feel that it added much information or was too vague to be of much value.

Fig. 1. Item flow of survey development.

Overarching themes from cognitive Interview Feedback

Learning more about pregnancy –

Women often commented on their desire and/or their attempts to learn more about their disability in the context of pregnancy that was often hard to obtain. Despite the general lack of information available to them, some women were confident in their pursuit of pregnancy while others expressed that their lack of knowledge of how pregnancy might affect their disability and specific risks made them reticent to consider pregnancy.

Influence of others –

The degree of influence of others (e.g., partners, family members, friends, and other loved ones) on a woman’s decision making process varied across respondents. Many expressed that they felt support from their loved ones, which kept them encouraged during their decision making process. Others explained that they had been discouraged by the people in their lives to pursue pregnancy; for some, this made them second-guess their decisions about pregnancy.

Preparing for pregnancy –

For a number of women, preparing for a pregnancy was often deeply contemplative and reflective, considering the anticipated changes in their lives during pregnancy and after giving birth. This typically preceded taking any concrete actions in planning, such as seeking more information from health care providers.

Comfort with healthcare providers –

Experiences with healthcare providers were often reflected upon in providing feedback on items. Some women had very positive interactions, feeling comfortable and trusting healthcare providers to guide them to a decision about pregnancy that was right for them. Others felt uncertain because health care providers never broached the topic with them or they seemed uncomfortable discussing pregnancy. Some were unsure how their providers could help them with their decision making about pregnancy.

Sub-section specific feedback

Experience of Making a Decision –

Common themes women thought about when answering questions in this section were the challenge of considering multiple trade-offs and risks; support systems and partnerships; having questions and needing information; knowing oneself and the desire to have children; knowing the potential impact of a pregnancy on their body; and sharing the decision with a partner. Women felt that nearly all questions in this section were important to ask. Importance was framed around concepts such as the need considering many aspects of a future pregnancy; need for support from others; understanding one’s emotions and thoughts about the decision; and the need for greater planning than women without disabilities may require.

Information about Disability and Pregnancy –

For this section, feedback on items were less in depth than other sections given the nature of the items. While not all topics were applicable to each respondent, when they were, women thought about risks associated with them (e.g., falling, bladder infections) and the importance of knowledge and information to manage them.

Things Affecting the Decision –

Common themes women thought about when answering questions in this section were caring for a baby; being prepared for unanticipated changes in their body and function; weighing risks to self and baby; managing the expectations of others; and needing knowledgeable health care providers. Women felt that nearly all questions were important. Importance was tied to understanding what is affecting their decision; knowing about risks that cause them worry; recognizing their connection to others (e.g., partners, family) in their decision; and having access to the right health care.

Knowing What is Important –

Common themes women thought about when answering were knowing oneself and the true desire for having children; being realistic; balancing independence and health with unknown changes during and after a pregnancy; and knowing a partner’s true wishes. Importance of the questions were tied to understanding the interaction of disability acceptance, wanting to have a baby, and accepting risk; understanding tolerance of changes in independence and health; and understanding how a partner’s wishes influences decision-making.

Support of Others –

Common themes were primarily around the different roles people would play in this decision; the boundaries of these relationships and their relative influence on the decision; and having differing levels of support from others. The importance of items were primarily around knowing who is and is not supportive and the greater need for support of women with physical disabilities than other women may need.

Communicating with Health Care Providers –

Common themes revolved around the relationship with and trust in health care providers; being understood; and the impact of knowledge gaps. Importance of items were tied to the need for health care providers to understand and support their patients; their partnership with patients; and women being empowered and a full participant in decision making.

Final survey content

The final version of the survey included six sub-sections and 114 items (see Fig. 1). We also included 2 open-ended items on what advice they would give to other women and health care providers about pregnancy. Survey data were collected from February through July 2019. The survey items, REDCap data dictionary and REDCap codebook, are given in Supplemental Materials.

Survey mode testing

Mode effects analysis focused on closed-ended items. Overall, the great majority of responses did not change (71.5–91.6%) or changed by only one response category (6.5–20.1%) between modes, whether the phone interview or the online survey was completed first. The averages for all change categories did not differ by which mode was completed first. There were no consistent patterns identified in the items that changed and brief follow-up interviews were conducted with participants to help clarify why changes occurred. These were generally due to the survey items prompting them to seek more information or think more about specific topics that the items addressed. Women generally preferred completing the survey online (See Supplemental Materials for more mode testing results.)

Assessment of survey performance

Missing Responses:

Across 47 closed-ended items in four sub-sections (experience of making a decision, things affecting a decision, knowing what is important and working with health care providers), missing responses were very low; the largest number of missing responses was 4 (3.5%) for three items.

Missing Content:

The majority of respondents did not have additional comments for the four sub-sections with that option (experience of making a decision, things affecting a decision, knowing what is important and working with health care providers; see Table 2). Based on comments, several themes were not well represented in the closed ended items: 1) discussion of alternatives to birth/pregnancy including vaginal birth vs C-section, difficulty of adoption, infertility, or the use of in vitro fertilization in women with disabilities; 2) if not having a child was discussed with them; 3) consideration of long-term effects of pregnancy and long-term care of child; 4) risk of screening for genetic factors to fetus; 5) the importance of ability to find an obstetrician with accessible examination table; and 6) if they had a full diagnosis or understanding of disability at time of pregnancy.

Not Applicable Responses:

The highest frequencies (>10%) of “not applicable” responses were for not passing on your genetic disorder to your child (19.3%; knowing what is important), coordinating personal care attendants (14.0%; knowing what is important) and the desire of your partner to have a genetic link to another person (13.2%; knowing what is important), talking to other women with physical disabilities who have been pregnant before (11.4%; things affecting the decision), and having enough emotional support from your partner (10.5%; experience of making a decision). Although well-reasoned a priori, in retrospect, the option of “not applicable” as a response option across all items within a sub-section (experience of making a decision, things affecting the decision and knowing what is important) resulted in a difficult-to-interpret response for some items where the applicability of “not applicable” was unclear.

Internal Consistency of Closed Item Sets:

The alpha values for the four sets of closed-ended items all indicated good to high internal consistency: experience of making a decision (0.886), things affecting the decision (0.779), knowing what is important (0.710), and working with health care providers (0.922).

Endorsement Frequency of Information about Pregnancy Items:

The endorsement of most items in the information about pregnancy sub-section and range of “other” responses suggest that the list is sufficiently comprehensive with the inclusion of 4 or 5 “other” options to capture a diversity of informational needs, see Table 4.

Table 4.

Endorsement of information about pregnancy items.

Topic	Endorsed (N, %)
Caring for an infant when you have a disability	88 (77.1%)
Medications you usually take (not related to fertility or pregnancy)	78 (68.4%)
Walking and/or ambulation	78 (68.4%)
Balance problems	70 (61.4%)
Spasticity or muscle tightness and cramping	66 (57.8%)
Other health problems, related or unrelated to your disability	66 (57.8%)
Self-care, like bathing, dressing or grooming	55 (48.2%)
Kidney or bladder infections	31 (27.1%)
Other 1	31 (27.1%)
Breathing difficulty	28 (24.5%)
Wheelchair fit, safety and/or maneuvering	25 (21.9%)
Transfers	25 (21.9%)
Other 2	12 (10.5%)
Other 3	8 (7.0%)
Other 4	3 (2.6%)
Other 5	0 (0%)

Discussion

The overall performance of the survey provides preliminary support for its content validity and utility to gather information about the pregnancy decision-making experience and informational needs of women with physical disabilities. Content validity of the new survey was supported in several ways. First, we involved women with physical disabilities as partners in all stages of its development; their perspectives were instrumental to ensuring that the content had sufficient breadth and depth to capture experience of respondents. Second, items were submitted to cognitive testing to gather feedback from the target population on interpretability and importance. Third, using the Ottawa Framework for Decision Support helped to identify concepts that were important to assess about pregnancy decision-making. Through several iterations, the final conceptual framework that underpins this new survey reflects key components of both the Ottawa model and lived experiences of women with physical disabilities.

Mode testing suggested that, generally, there were no effects on responding, which is consistent with other interview vs. online survey format examinations.²⁹ It also indicated that women generally preferred an online survey form. Unexpectedly, mode testing interviews also indicated that for some women, rather than a passive tool to gather information, the survey itself prompted them to reflect on their thoughts and feelings about pregnancy. When there was a discrepancy between modes, this tended to reflect a woman’s additional thought on the topic or search for more information after she completed the first round.

Although we did not design this survey to result in a single score or set of sub-scores, the acceptable to high internal consistency of closed-items within each survey sub-section shows promise for the survey’s further development towards generating interpretable scores and confirmatory factor analysis of these sub-sections. This can be very useful for comparing groups of women to identify differences in informational and decisional needs based on personal characteristics or situational factors. Currently, no survey tools exist to do this. Also, themes that can be considered in future iterations include discussion of alternatives to pregnancy and birth, long term effects of pregnancy and childcare, accessibility of physician offices, and risks of screening a fetus for disabilities.

One of the biggest challenges to evaluating the survey’s quality is that there are no consensus-based standards for evaluating item endorsement as a dimension of content validity or item coverage when methods such as factor analysis are not as useful for supporting structural validity. We also found it difficult to create items that mapped onto the construct of decisional conflict in the context of pregnancy decision-making. This may be due in part because this construct is less applicable when assessing past experience of decision-making, which was the case for many of our respondents. A particular strength of our approach to ensuring content validity of the new survey was involving stakeholders in the development of items and the overall structure of the survey and cognitively testing items with the target population.

While we designed this survey for women with physical disabilities, the content could be tailored for other types of disability or specific health conditions. It is our intention and hope that this survey, which was built on a template designed to be flexible, will serve as a foundation to explore the pregnancy decision-making experience of other cohorts of women. These data are critical to move the field towards interventions to support pregnancy decision-making in complex health contexts.