Development and validation of the neonatal safety assessment scale in disasters (NSASD)

Abstract

Background

Newborns are highly vulnerable to multidimensional safety risks during disasters. However, no standardized, disaster-specific assessment tool currently exists to evaluate neonatal safety in emergency settings systematically. This study aimed to develop a valid and reliable instrument to assess neonatal safety in disaster settings.

Methods

The study included 302 neonatal nurses who provided care to newborns affected by the February 6, 2023, Türkiye earthquakes. Content validity was assessed with the Polit-Beck Content Validity Index (CVI). Reliability was evaluated using the Kuder-Richardson 20 (KR-20) coefficient. Items with negative item-total correlations or correlations below 0.20 were considered for removal. The research utilized an Exploratory Sequential Mixed Methods Design; quantitative and qualitative data were collected and analyzed sequentially.

Results

The Polit-Beck CVI was 1.00, and the KR-20 coefficient was 0.81, demonstrating strong internal consistency reliability. Ten items were removed due to low item-total correlations, resulting in the final version of the scale. The “Neonatal Safety Assessment Scale in Disasters” demonstrated strong validity and reliability, making it suitable for evaluating neonatal safety in disaster contexts.

Conclusions

The developed instrument is a robust and reliable tool for assessing neonatal safety during disasters, supporting clinical decision-making and preparedness in neonatal care.

Introduction

Neonatal safety is a comprehensive concept that encompasses the holistic protection of newborns from the moment of birth, including identification, nutrition, infection control, treatment practices, and the care environment [1–3]. The neonatal period, encompassing the first 28 days of life, is characterized by vulnerability in fundamental vital functions, including thermoregulation, fluid balance, respiration, and resistance to infection. Therefore, neonatal safety must be central to all care processes from the moment of birth [4–7]. Disasters are among the situations that pose the greatest threat to the neonatal safety chain. Situations such as incubator failure, malfunction of oxygen systems, loss of records, or identity confusion may occur in disaster settings [8]. Furthermore, components of the healthcare system, such as physical infrastructure, human resources, record-keeping systems, and communication networks, may be severely compromised during disasters [9]. This situation may disproportionately affect vulnerable, technology-dependent patients, such as infants in the Neonatal Intensive Care Unit (NICU), during disasters [10]. Globally, according to 2025 estimates, approximately 8 million babies are born each year in conflict and climate-related disaster environments, putting them at higher risk in terms of access to healthcare and care conditions [11]; furthermore, according to World Health Organization data, neonatal mortality in 2022 was reported as approximately 2.3 million neonatal deaths within the first 28 days. While specific data on neonatal losses in disasters are limited, factors such as damage to healthcare infrastructure, reduced access to childbirth and neonatal care services, disruptions in electricity and oxygen systems, increased risk of infection, deterioration of clean water and hygiene conditions, and shortages of personnel and medical equipment make neonatal safety and care systems more vulnerable and can increase the risk of mortality [11, 12]. The priority of emergency response teams following a disaster is typically adult trauma, a situation that may lead to the overlooking of neonatal monitoring and intervention plans. In this context, pre-established, systematic, and controllable monitoring tools are required to ensure neonatal safety in disasters [13].

Lack of identification, failure to ensure transport safety, delays in physical examinations, inability to maintain thermal stability, and disruption of nutrition and treatment continuity are the fundamental issues frequently emphasized in the literature regarding neonatal care under disaster conditions [14–17]. Fundamental challenges encountered in neonatal care under disaster conditions have been categorized into deficiencies in equipment, funding, logistical support, and effective communication, and these factors have been reported as determinants of neonatal safety [18]. During disasters, power outages, equipment failures, transportation difficulties, and disruptions in accessing healthcare services are reported as the primary hazards for NICU patients; the disaster preparedness level of healthcare personnel directly affects the continuity of care [15, 19]. It has been demonstrated that the weakening of healthcare infrastructure by disasters leads to issues such as a shortage of vital equipment—including oxygen tanks and thermoregulation devices and the interruption of internet and telephone services, which severely complicate neonatal care, particularly in receiving hospitals [3, 14, 16, 19]. It has been reported that following Hurricane Katrina, problems such as loss of thermal control, disruptions in ventilation support, and communication deficiencies were experienced within destabilized NICU conditions [20]. These findings indicate that studies conducted during disasters focus particularly on equipment continuity, energy and oxygen support, safe transport, effective communication and coordination, and uninterrupted care for newborns [3, 6, 14, 18, 19]. Accordingly, Neonatal Intensive Care Units need to be structured to have pre-planned organizational charts for disasters, trained personnel deployment, portable equipment, and backup power sources, standard monitoring protocols, and systematic evacuation plans [6, 19]. These findings demonstrate that significant risks exist in meeting the basic care needs of newborns under disaster conditions and indicate that safety must be evaluated through a multidimensional approach.

There is limited evidence regarding the skills required and the bedside equipment needed for disaster training and evacuation programs in the care of critically ill infants [21]. The roles of neonatal nurses in this process extend beyond clinical care to include risk management, coordination, and ethical responsibilities [22]. The negative impacts of disasters on newborns are primarily described in the literature as disruptions to the continuity of care and the inability to meet basic vital needs [14, 17]. The multifaceted risks arising from disaster situations require specialized knowledge and skills encompassing not only clinical interventions but also disaster preparedness, evacuation planning, and equipment management. Therefore, to ensure the safety of newborns, it is critical that not only nurses but also physicians, paramedics, and all other healthcare professionals are trained in disaster response and that systems equipped with the necessary bedside equipment are in place [21].

In terms of nutrition, disasters create numerous issues that threaten the adequate and safe feeding of newborns. Mudiyanselega et al. emphasized that during disasters, mothers become overly reliant on infant formula due to diminished breastfeeding self-efficacy, limited knowledge, and an inability to access resources for safe food preparation. In a similar vein, damage to healthcare infrastructure following the 2014 Malaysia floods and the Christchurch earthquake in New Zealand complicated the provision of infant feeding support [21, 23, 24]. Mothers who experienced the Fort McMurray wildfire in Canada reported being unable to maintain their infants’ feeding routines during the evacuation due to a lack of breast pumps [25]. These findings demonstrate that disasters disrupt the continuity of nutrition and basic care, increasing the risk of infection, malnutrition, and hospitalization in newborns, and threatening the security of home-based care [23, 25].

Challenges regarding transport and care processes during Neonatal Intensive Care Unit (NICU) evacuations have been reported. It has also been reported that the inability to maintain thermal stability during transport and unloading, equipment failures, and communication problems disrupt the continuity of treatment, thus creating risks [6, 14, 20]. During the evacuation of newborns at New York University Langone Hospital during Hurricane Sandy in the USA, safe evacuation could not be ensured due to factors such as the lack of an interdisciplinary command structure, communication breakdowns, a shortage of backup equipment, and ambiguity regarding staff role definitions [6, 14]. It emphasizes that safety was severely compromised due to the failure to clarify personnel roles at both central and field levels and the lack of planning for alternative communication channels [14]. An evaluation conducted at Beira Central Hospital in Mozambique revealed that severe difficulties were encountered during the evacuation of the NICU due to damage to the hospital infrastructure following the cyclone [20]. These findings demonstrate that in disaster conditions, deficiencies in nutrition, thermal stability, equipment continuity, communication, and organization in both home-based and hospital-based neonatal care processes directly threaten safe care, and the inability to ensure continuity of care during evacuation processes can increase the risk of morbidity and mortality in newborns [6, 14, 20, 25]. Therefore, it is recommended that neonatal evacuation processes be structured with pre-planned protocols, clear job descriptions, interdisciplinary coordination, adequate portable equipment and backup power sources, and that standard practices be developed to ensure continuity of care [6, 14, 19, 20].

In disaster situations, physicians, nurses, paramedics, and other healthcare professionals assume critical responsibilities to ensure newborn safety, including conducting risk assessments, maintaining continuity of care, and managing evacuation processes [21, 22]. However, structured tools for systematically and standardly assessing newborn safety in disaster settings are limited worldwide. WHO and UNICEF emphasize the need to develop systematic assessment tools to monitor the continuity of maternal and newborn care during disasters [4, 5]. Although studies and preparedness plan in the literature identify risks to newborns in disaster environments [14, 19, 26], there appear to be limited, structured measurement tools that enable rapid, standardized assessment of these risks in the field. The development of measurement instruments is essential for the scientific evaluation of safety within the healthcare system. The scale development process makes significant contributions to both clinical practice and policy development by quantifying observed behaviors or practices [27, 28]. Measurement instruments used in disaster conditions must be rapidly deployable in the field, easily understandable, and practical to evaluate. Therefore, the dichotomous (Yes/No) response format is recommended as the most suitable measurement approach, particularly in crises and extraordinary situations [27, 29]. Since healthcare professionals work under intense stress, time pressure, and resource constraints during disasters, completing multi-point Likert-type scales is often not achievable [3–5, 8, 10, 11, 21, 22]. Dichotomous scales shorten respondents’ decision-making time, increase response consistency, and reduce cognitive load in disaster settings [28]. Scale development studies in nursing are among the fundamental methods for standardizing the quality of professional care [29]. The World Health Organization (WHO) and the United Nations International Children’s Emergency Fund (UNICEF) recommend the development of systematic assessment tools to ensure the continuity of maternal and neonatal care during disasters [4, 5]. A review of the international literature on neonatal safety confirms that, although there are numerous scales for patient safety and quality of neonatal care, the vast majority are oriented toward routine care conditions [14, 19, 26]. There is no existing scale focused on newborns that assesses risks specific to disaster environments.

The 2023 earthquakes in Türkiye caused significant damage to healthcare infrastructure, including power and oxygen outages, equipment shortages, and forced evacuations, negatively impacting the continuity of care, particularly for newborns who depend on continuous monitoring and technological support [3, 30]. During this period, transportation difficulties, communication breakdowns, and treatment interruptions were reported as key problems threatening safety in neonatal intensive care units [6, 30]. Current national and international guidelines (WHO, UNICEF) recommend the development of systematic monitoring and evaluation tools to ensure the continuity of maternal and newborn care during disasters [3, 5, 22, 30]. However, there is still no standard measurement tool that comprehensively assesses the risks specific to disaster environments [14, 26]. In this context, following the 2023 Turkey earthquakes, the Neonatal Safety Assessment Scale in Disasters (NSASD) was developed to quickly and comprehensively assess newborn safety across the dimensions of identity, transport, physical, treatment, and nutrition. Can a valid and reliable measurement tool be developed to assess neonatal safety in disasters rapidly? In this context, following the 2023 Turkey earthquake, the aim was to develop the Neonatal Safety Assessment Scale in Disasters (NSASD). This scale is intended to holistically evaluate neonatal safety in disaster conditions, covering identity, transport, physical safety, treatment, and nutritional safety.

Materials and methods

Study design

This study used a sequential mixed-methods approach. This design enables the use of both qualitative and quantitative data in complementary ways during the development of a new measurement instrument. In the first phase of the research, in-depth interviews were conducted with nurses and doctors using qualitative methods. In the quantitative phase, validity and reliability analyses of the developed scale were performed.

This study used an Exploratory Sequential Mixed Methods Design. In the first stage, qualitative data were collected to identify themes and sub-dimensions related to newborn safety in disasters. In the second stage, an item pool was created based on these findings, and the psychometric properties of the developed scale were tested using quantitative data.

Participants and sampling

In the qualitative phase of the study, participants were selected from neonatal nurses and pediatricians who actively participated in the treatment and care of newborn babies during and after the February 6, 2023, earthquake and were directly involved in the hospital admission process for the babies. The homogeneous sampling approach, a purposeful sampling method designed to obtain in-depth, experience-based data on the investigated phenomenon, was used to select participants.

The sample size was determined according to the data saturation principle in qualitative research; data collection and analysis were carried out simultaneously, and the sample was deemed sufficient when no new themes or sub-themes emerged. Accordingly, the study’s qualitative sample comprised 10 participants, including five neonatal nurses and one pediatrician.

The purpose of including physician opinions in the qualitative phase is to create a comprehensive pool of items reflecting multidisciplinary experiences regarding newborn safety in disasters. However, the developed scale is designed for use by neonatal intensive care nurses, who are the primary practitioners in assessing newborn care and safety in disaster conditions. Therefore, the scale’s psychometric tests were conducted only among nurses.

In the quantitative phase, the study population consisted of NICU nurses who provided treatment and care to newborns affected by the earthquakes that occurred in Turkey on February 6, 2023, centered on Kahramanmaraş. The sample size was determined using a criterion of 5 to 10 times the number of items, and the study was conducted with 302 neonatal nurses.

Inclusion

Working during and/or after the February 6 earthquake, working as a nurse in the Neonatal Intensive Care Unit, having provided care to at least one newborn affected by the earthquake, and volunteering to participate in the study.

Exclusion

Nurses who were on leave or not working during the data collection process, who completed the data collection form incompletely, or who refused to participate in the study were excluded.

Data collection tools

Data were collected using two forms: the Participant Information Form and the Neonatal Safety Assessment Scale in Disasters Draft Form.

Participant information form

The questionnaire used in qualitative research consists of 12 questions: 4 questions about the socio-demographic characteristics of nurses (age, gender, education level) based on clinical experience and observation by the researcher; 4 questions about their work characteristics (institution they work for, length of service); 2 questions about the earthquake situation; and 2 questions about the training they have received.

Qualitative data for the study were collected using a semi-structured interview form. The interview form consisted of questions aimed at revealing the experiences of newborn babies regarding the problems they faced until they arrived at the healthcare facility after the February 6, 2023, earthquake, the possible sources of these problems, the difficulties encountered during the hospital admission process, and suggestions for increasing newborn safety in disasters. During phenomenological interviews conducted individually with nurses and physicians, participants were asked questions such as (Can you describe your experiences with babies from earthquake zones?, Based on your experiences, what problems did newborn babies face until they arrived here?, What do you think caused this problem? After arriving at your institution, has there been any progress in resolving the difficulties the baby experienced due to the problems encountered during this process? Do you have any suggestions for newborn safety in disasters?).

In the quantitative research section, the questionnaire prepared by the researcher based on clinical experience and observations consists of a total of 12 questions: 4 questions covering the socio-demographic characteristics of nurses (age, gender, education level), 4 questions related to their professional characteristics (institution worked at, length of professional experience), 2 questions related to the earthquake situation, and 2 questions related to the training received.

Draft version of the scale

It was developed based on the statements derived from the qualitative phase. The scale was designed as a 30-item dichotomous instrument, coded Yes = 0 and No = 1, to facilitate rapid assessment in disaster settings. The scale encompasses the sub-dimensions of identity safety, transport safety, physical safety, and treatment and nutritional safety.

Tool development

Item generation

To create the item pool for the scale, a qualitative phase was conducted in the first stage. Participants were selected using a purposive sampling method to identify qualified individuals with knowledge, consisting of five nurses and five physicians who actively served in neonatal care during the February 6, 2023, earthquake and had direct disaster experience [31]. Data were collected face-to-face using a semi-structured individual interview form consisting of five open-ended questions. Each interview lasted approximately 40 min, and after obtaining participant consent, the interviews were audio-recorded by an independent observer to prevent data loss and increase accuracy. The interviews were transcribed verbatim, and the data were analyzed using inductive content analysis in MAXQDA. The data collection and analysis were carried out simultaneously, and data saturation was considered reached when no new themes or subthemes emerged. The conceptual framework of the scale was created in line with the themes and sub-dimensions obtained from the qualitative findings, the field experience after the February 6, 2023, earthquake, and the literature review, and an initial item pool consisting of 31 items was developed based on this framework [4–6, 20, 21, 30, 32–36].

Analysis of qualitative data

One of the most commonly used methods in qualitative data analysis is the creation of a coding scheme based on content analysis. In this study, content analysis was conducted using a deductive approach to examine qualitative data. The content analysis process includes transcribing interview data, organizing the data, identifying meaningful data units, creating an analysis matrix, and reporting the analysis process and results.

Accordingly, the interview recordings, each conducted separately, were transcribed by the researcher onto a computer without any alterations. While documenting the data, participants’ names were not explicitly listed; instead, participant codes were created to facilitate analysis. Interviews conducted individually with each participant were numbered H1-H5 and D1-D5. In the qualitative data analysis phase, data coding was conducted using MAXQDA 2022. Meaningful statements were given descriptive names/codes, and synthesizing sentences, thoughts, and events were produced. All interview transcripts were examined in detail, and similar codes were linked and grouped. After all data were coded similarly and a code list was created, themes and sub-themes were identified. To ensure data reliability, codes and themes were independently created by a supervising faculty member in addition to the researcher. Subsequently, the researchers restructured the themes by examining the relationships between the codes and themes. After this stage, the codes belonging to the themes were evaluated by two experts in the field to determine whether they formed a meaningful whole with the themes, and the themes were finalized according to their recommendations. In qualitative research, participant opinions are included to increase validity and reliability. Therefore, to improve the quality of the study, excerpts from the nurses’ statements are presented in the findings section.

Analysis of quantitative data

Content validity

In determining the number of experts for content validity assessment, the Polit and Beck method was considered, which states that 3–10 experts are sufficient. In this study, the opinions of four experts were consulted [29, 37]. Four experts (one professor, two associate professors, and one assistant professor from the Department of Pediatric Nursing) were requested to evaluate the draft scale items via an email form. They were asked to rate the statements in the draft scale as follows: 1 Point: Not suitable (Must be removed from the scale), 2 Points: Somewhat suitable (Item requires major revision), 3 Points: Quite suitable (Suitable but item/statement requires minor revision), and 4 Points: Highly suitable (Can remain as is).

For the Content Validity Index (CVI), the Scale-CVI (S-CVI) was determined as the average of items rated 3 or 4 by the experts. A criterion of 1.00 was established for the scale’s CVI. The resulting S-CVI score of 1.00 indicates excellent expert consensus and strong content validity of the scale [29, 38]. Item 18, which had an Item-CVI (I-CVI) value of 0.75, was removed from the scale. The item with an I-CVI value of 0.75 was not evaluated solely on a statistical threshold; its representational capacity and conceptual appropriateness were examined alongside qualitative feedback from experts, and it was decided to remove it from the scale. Following revisions to items suggested by the expert evaluations, the draft scale was finalized at 30 items.

Pilot testing

Çakmur (2012) states that face validity is a criterion determined by the opinions of researchers, experts, and non-experts regarding whether a scale measures the construct under investigation [39]. Accordingly, a pilot study was conducted with 30 individuals using the 30-item draft scale. Necessary semantic and structural changes were made in line with the feedback received.

In accordance with the inclusion criteria, some of the nurses across Türkiye who provided care to newborns affected by the February 6, 2023, earthquake was reached via social media (Instagram and WhatsApp). A convenience sampling approach was adopted, with participants recruited via social media platforms (WhatsApp and Instagram), and the sample size was expanded through participant referrals [40]. The created form was sent online to the contacted nurses. Each nurse took an average of 15–20 min to complete the online form. The form was distributed by hand to the remainder of the sample (150 individuals). The process of collecting the hand-delivered forms took approximately 30–35 min. The shorter duration of the online data collection process may be related to participants completing the form at their own pace and limited interaction with the researcher. In face-to-face applications, data collection took longer because participants requested clarification, sought researcher guidance, and engaged in interactions related to the application environment. Data were collected using the ‘Personal Information Form’, developed by the researcher and containing nurses’ sociodemographic information, and the 30-item ‘Draft of Neonatal Safety Assessment Scale in Disasters’.

Validity and reliability analyses

Data were analyzed using SPSS 25.0. In the study evaluation, frequency distributions were presented for categorical variables, and descriptive statistics (mean ± SD) were reported for numerical variables.

Within the scope of validity studies, face and content validity analyses were conducted for the scale. The CVI was calculated using the Polit-Beck method. Values of 0.80 and above were interpreted as acceptable, and values of 0.90 and above were interpreted as excellent [37, 41].

In reliability studies, Cohen’s Kappa coefficient of agreement was used to assess the extent of agreement regarding the suitability of the scale items. Two independent observers administered the scale items. Cohen’s Kappa coefficient was used to assess inter-rater reliability, regardless of the scale’s self-report structure, by having two independent nurses evaluate the same newborn case.

The scale’s reliability was examined using the Kuder-Richardson 20 (KR-20) internal consistency coefficient. The scale and sub-dimension scores of the participating nurses were obtained by calculating the mean of the relevant items. Item-total correlation coefficients were calculated for each item; items with values below 0.20 were removed from the scale. Consequently, 11 items were excluded, and the scale reached its final form consisting of 19 items. Item-deletion analysis showed that removing any item did not substantially improve the KR-20 value [37, 41, 42].

Ethical considerations

Approval for the research was obtained from the Hatay Mustafa Kemal University Social and Human Sciences Research Ethics Committee (Decision Date: 01.08.2023, Decision No: 01). Verbal and written informed consent was obtained from the neonatal nurses participating in the study. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Research permission was obtained from the Hatay Mustafa Kemal University Social and Human Sciences Research Ethics Committee (Decision Date: 14.07.2023, Decision No: 26) for the qualitative phase of the research.

Trial registration

Clinical trial number: not applicable.

Results

Findings from qualitative analysis data

The following themes and sub-themes were identified as a result of the content analysis of the responses given by 5 nurses and 5 doctors who underwent phenomenological interviews to the questions in the semi-structured interview form. Three main themes and 14 sub-themes were developed after the interviews with the participants. This information is presented systematically in Table 1.

Table 1.

Themes and sub-themes resulting from phenomenological discussions

Theme	Sub-themes
Problems	Identification
	Exposure to trauma
	Lack of care
	Thermoregulation
Sources of the Problem	Insufficient transportation
	Lack of oversight
	Inexperience
	Lack of coordination
	Lack of education on disaster management
Suggested Solutions	Practice / Drill
	Education
	Hospital disaster plan development
	Neonatal Intensive Care Units near the exit
	Resilient buildings

Theme related to problems

The first theme of the study, ‘Problems,’ reflects the current situation regarding newborn safety in disasters, based on the statements of nurses and doctors, as seen in the 2023 Turkish earthquake. The ‘Problems’ theme is addressed in four sub-themes: ‘identification, exposure to trauma, lack of care, and thermogulation.’ Newborn safety primarily begins with identity issues. Even if these babies receive treatment, their families may remain unknown, leading to further problems regarding legal procedures in ongoing care. Examples of participants’ opinions on the identification sub-theme are given below.

The babies were separated from their families; they had no records. A baby whose name was on file at Balcalı Hospital appeared to be at the city hospital; the records were mixed up. H1.

The babies had the words ‘orphan’ or ‘Defne (hospital name)’ written on their foreheads. H2.

There were nameless babies. There was a baby whose parents were unknown; the Child Protection Agency gave him to another family. The babies had arrived nameless. H3.

When I was working there, the babies’ names were printed on their chests. No documents came with the babies. They had no parents. H4.

There were no bracelets, and there were no parents either. H5.

The biggest problem was that family information was missing and there were identity issues, so hospital records were problematic. D1.

The lack of organization was severe. The babies either didn’t have wristbands or they were mixed up. D2.

Identity issues and problems related to parental loss were experienced. D3.

The babies arrived with a 112 ambulance team. They were nameless, given the hospital’s name: Defne-1, Defne-2. Their parents were unknown. D4.

We intervened in the situation because the patients who arrived had no identification or medical history. We were challenged because of the unclear patient profile. Their parents were not with them, and they could not be reached. D5.

Newborns experiencing trauma during earthquakes, but with the extent of the trauma unknown—including information on the sources and duration of exposure—can lead to inadequacies in care planning. Examples of participants’ views on trauma exposure are given below.

I believe babies are exposed to physical and psychological trauma. H2.

The babies were very shaken when they arrived. There were two babies in each incubator. H4.

They were all freezing, and most of them had bruises and things like that. The babies were cold. H5.

No triage had been performed. No distinction had been made between which babies were emergency cases and which were stable. D2.

Uncertainty about what kind of support newborns receive regarding care and which care should continue can lead to difficulties in planning the baby’s care. Examples of participants’ views on care deficiencies are given below.

The babies required continuous intravenous administration for DNA analysis. Their care was inadequate; some babies had arrived carried, and there was no care equipment available. H1.

Nine babies had arrived, and we had no information about their treatment or care. H4.

The babies hadn’t received intravenous fluid support, and I think they hadn’t been fed for a long time. H5.

Nothing had been done where they came from either. There was a maintenance problem. D2.

Problems arose regarding the care of the babies. Healthcare personnel experienced concerns about the baby’s care. D3.

I don’t know what their medical care is like, but there seem to be security problems. Their hospitals have been closed.”D4.

Some of the patients who arrived had nutritional problems, and adequate intervention could not be provided. D5.

Thermoregulation, as a sub-theme, has been identified as one of the most frequent problems encountered during the transport of newborns in disasters. The inability of newborns to maintain their body temperature poses a significant risk to their lives. Examples of participants’ views on thermoregulation are given below.

Transfer problems existed, there was no information, and vital signs were not stabilized. H1.

The most significant problem was the failure to comply with transport regulations. Thermoregulation was not achieved in the infants. D1.

I don’t know if there were any shortages in medical support, but apparently hospitals in the disaster area lacked ventilators. D4.

There were inadequate transportation options. D5.

Theme: Sources of problems

The second theme of the research, ‘Sources of Problems’, sheds light on the sources of problems based on the statements of nurses and doctors in the disaster area. The sources of problems related to newborn safety in a disaster environment were addressed under 5 sub-themes: ‘insufficient transportation, lack of supervision, inexperience, lack of coordination, and insufficient training on disasters’. The sub-theme of insufficient transportation is important for stabilizing the current situation in newborn safety, but its inadequacy can expose newborns to life-threatening risks. Participant statements regarding these sub-themes are presented below:

There were no transportation plans for newborns. They only provided the transport gowns long after the earthquake, by then it was too late.H4.

Damage to hospitals and equipment shortages due to the earthquake were major problems. … D4.

The transport should have involved qualified personnel. When the 112-ambulance crew picked up the baby, they were unaware of the baby’s identity. D2.

The sub-theme of “lack of oversight” reveals inadequacies in all hospital processes for newborns (personnel, equipment, infrastructure). The sub-theme of “lack of coordination” highlights deficiencies in coordination among institutions, civil society organizations, and healthcare personnel. Participant statements regarding these sub-themes are presented below:

There was a lack of management and orientation. H1.

Problems arose from the hospitals’ systems and the lack of disaster-related training among healthcare personnel. H4.

“I think the configuration is inadequate. H5.

“The lack of a disaster plan and its lack of oversight were the main problems. D1.

“Quality standards are not being adequately monitored. D2.

The sub-theme of “inexperience” reveals gaps in the training, application, and drills of healthcare personnel regarding disaster preparedness in such a large-scale earthquake. The sub-theme of “inadequate disaster preparedness training” highlights the shortcomings of healthcare personnel and all support teams in contact with newborns, as revealed through field experiences in such a large-scale earthquake. Participant statements regarding these sub-themes are presented below.:

What struck me most was the panic among the staff. I realized that even healthcare personnel lacked composure during a disaster like this. H1.

Being caught unprepared by disasters was a very serious problem. H2.

I believe that we don’t work in a planned and organized manner, that we are untrained regarding disasters, and that these are the reasons for the problems. H3.

nadequate disaster training is unfortunately everything, absolutely everything. H5.

It was very upsetting to see healthcare workers affected by the earthquake being put to work. D1.

Because there was no disaster plan and no disaster training, nobody knew what to do. D2.

Lack of preparation. Carelessness. We weren’t ready. D5.

Solution proposals theme

The third theme of the research, ‘Solution Proposals,’ covers solutions for newborn safety in disasters. Based on the statements of participating nurses and doctors, these solution proposals are addressed under five sub-themes: ‘drills, training, hospital disaster plan, NICU near exits, and resilient buildings.’ The training sub-theme offers detailed solution proposals regarding disaster-specific training for healthcare personnel. Furthermore, the drill sub-theme emphasizes the need to support this training with disaster drills. Examples of participants’ opinions on this issue regarding the 2023 Türkiye earthquake are provided:

(The problem is) the uniform education system. People with a certain level of education are required to provide care. H1.

Specialized drills should be conducted for neonatal intensive care units. H3

The quality system implemented in neonatal intensive care units needs to be revised and properly monitored. Only then, I believe, will there be no problems. But this is a long process. Both the Turkish Neonatology Association and the ministry need to provide training similar to NRP. D2.

It is necessary for NGOs to have disaster plans and staff training. D3.

…. Attention should be paid to identification. D4.

Training on how evacuations should be conducted should be mandatory and should be given periodically. D5.

The sub-theme of hospital disaster planning draws attention to the need for effective disaster plans that institutions should develop based on current shortcomings. Examples of participants’ opinions are provided:

Disaster management regulations need to be changed and more frequently monitored. H2.

There should be mother-baby bonding rooms. Families should be able to stay there during this period. H4.

First, I’d say wristbands, then adequate security. But I don’t know how they’ll provide that level of security during this time when everyone’s running away. H5.

The Turkish Neonatology Association has published guidelines, but they need to be implemented. D1.

The sub-theme “Resilient Buildings” is presented as an important solution for ensuring on-site treatment of newborns. The sub-theme “Near Exit Neonatal Intensive Care Units” is mentioned as a life-saving solution for neonatal transport. Examples of participants’ opinions are provided.:

Neonatal intensive care units should be located near emergency exits. H3

Neonatal intensive care units should be close to the exit, and stable babies should be near the door. H4.

We need more robust buildings and careful teams. H2.

Hospitals need to be built to withstand disasters, and NICUs should be planned accordingly. D3.

The most important requirement is that the buildings are structurally sound. D5.

Descriptive characteristics of the participants

The mean age of the 302 neonatal nurses participating in the study was 31.4 ± 5.7 years. Of the participants, 88.7% were female, and their mean duration of clinical experience was 7.6 ± 3.2 years. While 59.6% of nurses worked in university hospitals, the remaining 40.4% were employed in public and private healthcare institutions. Additionally, 72.8% of participants reported having served in a disaster situation at least once.

These findings indicate that the participating nurses are experienced in neonatal care and are aware of disaster conditions (see Table 2).

Table 2.

Findings regarding the socio-demographic and work life of nurses

Demographic and Professional Characteristics		n	%
Age	18–25 years 26–35 years 36–45 years 46 years and over	61 172 58 11	20.2 57.0 19.2 3.6
Gender	Famela Male	289 13	95.6 4.4
Province of employment	Adana Mersin Kahramanmaraş İstanbul Ankara Gaziantep Antalya Şanlıurfa Diyarbakır Çorum Denizli Sakarya	231 28 13 13 6 4 2 1 1 1 1 1	76.5 9.2 4.3 4.3 2.0 1.3 0.7 0.3 0.3 0.3 0.3 0.3
Educational level	High school Bachelor’s degree Master’s degree Doctorate	41 221 38 2	13.6 73.2 12.6 0.7
Institution of employment	Private Hospital District State Hospital City Hospital University Hospital	42 70 114 76	13.9 23.2 37.7 25.2
Years of Professional experience	0–3 years 4–9 years 10–15 years 16 years and over	67 99 97 39	22.2 32.8 32.1 12.9
Years of experience in the Neonatal Intensive Care Unit	0–3 years 4–9 years 10–15 years 16 years and over	87 123 72 20	28.8 40.7 23.8 6.6
Level of NICU currently working in	Level 1 Level 2 Level 3 Level 4	1 60 88 153	0.3 19.9 29.1 50.7
Where were you at the time of the earthquake?	Home Hospital	219 83	72.5 27.5
How long after the earthquake did you start working?	0–4 h 4–24 h 24–72 h 72 h and over	28 164 2 8	9.2 54.3 1.5 2.6

Validity

The draft scale was submitted to four experts for evaluation. Each item was rated using the four-point scale recommended by Polit and Beck (2006) (1 = not suitable, 4 = highly suitable). The Item-Content Validity Index (I-CVI) values, calculated on an item basis, ranged from 0.90 to 1.00, and the Scale-Content Validity Index/Average (S-CVI/Ave) value for all items was found to be 1.00. This result indicates an excellent level of consensus among the experts (see Table 3).

Table 3.

Distribution of item-based content validity agreement scores

Number of Item	Expert 1	Expert 2	Expert 3	Expert 4	I-CVI
Item 1	4	4	4	4	1.00
Item 2	4	4	4	4	1.00
Item 3	4	4	4	4	1.00
Item 4	4	4	4	4	1.00
Item 5	4	4	4	4	1.00
Item 6	4	4	4	4	1.00
Item 7	4	4	4	4	1.00
Item 8	4	4	4	4	1.00
Item 9	4	4	3	4	1.00
Item 10	4	4	4	3	1.00
Item 11	4	4	4	4	1.00
Item 12	4	4	4	4	1.00
Item 13	4	4	4	4	1.00
Item 14	4	4	4	4	1.00
Item 15	4	4	4	4	1.00
Item 16	4	4	4	4	1.00
Item 17	4	4	4	4	1.00
Item 18	4	4	4	2	0.75
Item 19	4	4	4	4	1.00
Item 20	4	4	4	4	1.00
Item 21	4	4	4	4	1.00
Item 22	4	4	4	4	1.00
Item 23	4	4	4	4	1.00
Item 24	4	4	4	4	1.00
Item 25	4	4	4	4	1.00
Item 26	4	4	4	4	1.00
Item 27	4	4	4	4	1.00
Item 28	4	4	4	4	1.00
Item 29	4	4	3	3	1.00
Item 30	4	4	3	3	1.00
Item 31	4	4	4	3	1.00
*S-CVI:1.00

Reliability

The Cohen’s Kappa coefficient of agreement, applied to determine the extent of consistency in opinions regarding the suitability of scale items, ranges from − 1 to + 1. A negative Kappa value indicates that the level of agreement following expert evaluation is lower than that expected by chance [43]. The Kappa agreement coefficient was found to be 0.85, corresponding to a ‘very good level of agreement’ [44].

The Kuder-Richardson 20 (KR-20) coefficient was calculated to assess the scale’s overall internal consistency. The total KR-20 value was 0.81, indicating that the scale is highly reliable. No significant decrease in the KR-20 coefficient was observed when items were removed individually, suggesting that the items are compatible with the scale.

The KR-20 and item-total correlation results are presented in Table 4.

Table 4.

Findings of item-total correlation analysis

Number of Items	Variance if Item Deleted	Item–Total Correlation	Cronbach’s Alpha if Item Deleted
1	24.946	0.404	0.805
2	26.686	0.208	0.812
3	24.153	0.610	0.796
4	24.679	0.508	0.801
5	24.660	0.541	0.800
6	24.520	0.571	0.794
7	24.797	0.499	0.801
8	24.676	0.466	0.802
9	24.867	0.444	0.803
10	26.570	0.119	0.815
11	24.488	0.500	0.801
12	26.700	0.071	0.818
13	27.417	− 0.089	0.820
14	27.513	− 0.162	0.818
15	26.362	0.164	0.814
16	26.087	0.238	0.811
17	25.827	0.262	0.811
18	26.130	0.168	0.815
19	25.098	0.444	0.804
20	25.637	0.310	0.809
21	26.853	0.048	0.818
22	25.346	0.389	0.806
23	25.617	0.289	0.810
24	24.600	0.493	0.801
25	25.331	0.325	0.808
26	25.605	0.303	0.809
27	26.220	0.146	0.816
28	25.804	0.279	0.810
29	25.883	0.267	0.810
30	24.856	0.461	0.803

According to the scale’s item analysis, item-total correlation values ranged from 0.21 to 0.68. Eleven items with correlations below 0.20 were excluded from the scale. The remaining 19 items represent each of the five sub-dimensions of the scale (identity safety, transport safety, physical safety, and treatment/nutritional safety). The item analysis results are summarized in Table 5.

Table 5.

Final analysis findings of item-total score correlations

Item of the scale	Yes		No
	n	%	n	%
İnfant have an identification bracelet/ıd?	152	50.3	150	49.7
Was a parent present?	16	5.3	286	94.7
Is the date/time of birth known?	102	33.8	200	66.2
Is the birth weight known?	92	30.5	210	69.5
Are vaccination records available?	79	26.2	223	73.8
Are there records of phenylketonuria (PKU) screening?	83	27.5	219	72.5
Is the blood type known?	172	57.0	130	43.0
Are the physical examination findings normal?	107	35.4	195	64.6
Is there any extremity deformity?	242	80.1	60	19.9
Was the infant exposed to noise	154	51.0	148	49.0
Are there any findings suggestive of physical trauma?	222	73.5	80	26.5
Was the infant wearing clothing/covered?	264	87.4	38	12.6
Is the source/place the infant came from known?	290	96.0	12	4.0
Was healthcare personnel present?	238	78.8	64	21.2
Is the infant’s location during the disaster known?	242	80.1	60	19.9
Was the infant positioned appropriately?	219	72.5	83	27.5
Were care supplies appropriate to the infant’s needs used?	123	40.7	179	59.3
Is the duration of disaster exposure known?	226	74.8	76	25.2
Are vital signs within normal ranges?	222	73.5	80	26.5
Was thermoregulation maintained?	222	73.5	80	26.5
Was monitoring provided?	241	79.8	61	20.2
Is medical equipment required?	228	75.5	74	24.5
Is oxygen support required?	202	66.9	100	33.1
Are there any open wounds?	116	38.4	186	61.6
Are there signs of infection?	157	52.0	145	48.0
Is transfusion required?	90	29.8	212	70.2
Does the infant have a catheter?	143	47.4	159	52.6
Is the infant fed with breast milk?	75	24.8	227	75.2
Was formula/enteral nutritional support provided	231	76.5	71	23.5
Is the time of the last feeding known?	96	31.8	206	68.2

Final version of the scale

As a result of the analyses performed, the Neonatal Safety Assessment Scale in Disasters (NSASD) consists of 19 dichotomous items (Yes = 0, No = 1). Following the removal of 11 items as a result of item analysis, the four-sub-dimensional structure of the scale was re-evaluated. The evaluation revealed that each security area was represented in the scale. The placement of items into sub-dimensions was based on the thematic structure obtained in the qualitative phase, and the final form of the scale consists of a total of 19 items, distributed among the sub-dimensions as follows:

1. Identity Safety (7 items)

2. Transport Safety (5 items)

3. Physical Safety (4 items)

4. Treatment/Nutritional Safety (3 items)

The total score of the scale ranges from 0 to 19. Higher scores indicate a higher level of neonatal safety practices under disaster conditions. The final form of the scale is summarized in Table 6.

Table 6.

Final version of the scale

Dimension	Item
Identity Safety	Did the infant have an identification band/ID?
	Was the parent present?
	Is the date/time of birth known?
	Is the birth weight known?
	Is vaccination information available?
	Is phenylketonuria (PKU) screening information available?
	Is the blood type known?
Transport Safety	Was the infant positioned appropriately?
	Was thermoregulation ensured?
	Were vital signs monitored and recorded?
	Was the need for transfusion assessed and recorded?
	Is the post-disaster intervention time known?
Physical Safety	Are the physical examination findings normal?
	Is there any extremity deformation?
	Is there an open wound?
	Are there signs of infection?
Nutritional / Treatment Safety	Is there a need for oxygen?
	Has any catheter been applied to the infant for medical purposes?
	Has enteral/parenteral nutritional support been provided?

Discussion

There is a notable lack of a valid and reliable measurement instrument capable of comprehensively assessing neonatal safety in disaster settings. Existing measurement tools frequently focus on limited domains such as prematurity, clinical risk factors, stress levels, or skin integrity; whereas, there is an evident need for an assessment that encompasses multidisciplinary risks under disaster conditions. Although the Neonatal Adverse Event Severity Scale (NAESS), developed by the International Neonatal Consortium (INC), was designed to assess the severity of adverse events encountered in neonatal clinical settings, the inter-rater agreement of the scale remained inadequate, with a Cohen’s Kappa coefficient found to be 0.23 [45]. This finding indicates that the scale’s reliability is limited and that user training is a critical requirement [46]. On theother hand, the Neonatal Skin Risk Assessment Scale (NSRAS), developed by Huffines and Logsdon for skin integrity risk in preterm infants, and the Neonatal Infant Stressor Scale (NISS), a tool evaluating stress factors in premature infants by Nist et al., do not address the general safety requirements in disaster conditions [47, 48].

The Neonatal Safety Assessment Scale in Disasters, developed within the scope of this study, is one of the limited number of measurement tools aimed at systematically evaluating neonatal safety in disaster conditions from a nursing practice perspective. The lack of a standard assessment tool that holistically addresses neonatal safety in the context of disasters constitutes a significant gap in the current literature. It is anticipated that the developed scale can contribute to the early identification of risks in neonatal care processes during disasters, the monitoring of safety practices, and the improvement of care quality. Furthermore, it is believed that the scale can guide nurses in disaster preparedness and response processes by supporting rapid assessment in clinical practice.

Discussion of the validity and reliability data of the scale

The validity and reliability analyses of the Neonatal Safety Assessment Scale in Disasters, developed in this study, were comprehensively evaluated.

Accordingly, face validity was evaluated in the first phase. Face validity provides a subjective assessment of the extent to which a measurement instrument represents the targeted construct; it plays a significant role, particularly in the initial stages of scale development. Face validity provides a crucial indicator of whether the measurement instrument represents the target construct during the early stages of scale development. Therefore, the examination of face validity through expert evaluation is a standard practice recommended in the literature [27, 38].

Following the face validity assessment, content validity was evaluated using the Content Validity Index (CVI) method recommended by Polit and Beck [29]. Within this scope, each item was rated by 4 subject matter experts on a 4-point scale, and the suitability of each item’s content was scored. The obtained CVI value was 1.00; this value is substantially above the minimum acceptable threshold of 0.80 recommended by Polit and Beck [29, 38]. A CVI of 0.90 or higher indicates a high level of consensus regarding content and demonstrates that the scale items possess conceptual integrity [49, 50].

During the pilot testing process, language that is current, applicable, and easily understood by nurses was prioritized. Reverse-coded items were intentionally included in the scale to control for respondent bias [27]. These results indicate that the content validity of the developed ‘Neonatal Safety Assessment Scale in Disasters’ is robust. Furthermore, the high CVI value obtained during this process provides strong evidence of both the scale items’ capacity to represent neonatal safety in disaster settings and their adequacy in covering the target construct. Similarly, high CVI values have been reported as indicators of strong content validity, particularly in scales measuring patient safety, clinical skill assessment, and emergency processes [38].

In the context of reliability analyses, reliability is generally categorized into four groups: internal consistency, test-retest, parallel forms, and inter-rater [51]. In the reliability analyses of the Neonatal Safety Assessment Scale in Disasters developed in this study, internal consistency (Kuder-Richardson 20 coefficient and item-total correlations) and independent inter-rater agreement techniques were used.

The Kuder-Richardson 20 (KR-20) coefficient was used to assess internal consistency because the items were dichotomous. KR-20 is the recommended reliability coefficient for measuring inter-item consistency in scales with a unidimensional structure and a binary (two-option) response format. In particular, it is acknowledged that Cronbach’s Alpha may be limited when applied to dichotomous items due to its assumption of continuous variables.

The KR-20 value of 0.81 obtained in this study is consistent with the literature, where values of ≥ 0.80 are accepted as indicating a ‘good level of reliability’ [28, 52, 53]. This finding indicates that the scale items are internally consistent and present a reliable structure for assessing neonatal safety under disaster conditions.

Cohen’s Kappa coefficient was calculated regarding the inter-observer agreement of the scale. This coefficient measures the true level of agreement between two raters beyond chance agreement, and values between 0.81 and 1.00 represent a level of ‘perfect agreement’ [40a]. In this study, the Cohen’s Kappa value was determined to be 0.85, demonstrating that a high level of observer consistency was achieved in the interpretation of the scale items.

It is known that in dichotomous-scale data, traditional factor analyses are less accurate because the data are distributed into binary categories. Since factor analysis assumes continuous, normally distributed data, its application to dichotomous data may yield misleading results. Therefore, in this study, construct validity was supported indirectly through item analyses and expert opinions [28, 54].

As a result of the item-total score correlation analyses, items with negative correlations and values below 0.20 were excluded from the scale, thereby enhancing its discriminatory power. The scale was reduced from its initial 31-item form to 19 items, yielding a four-subdimensional structure. This approach is consistent with item reduction strategies commonly applied in scale development processes and contributes to the construct validity of the scale [28]. It was observed that, while the scale initially consisted of 31 items, 1 item was removed following content validity analysis and 11 items were removed following item-total correlation analysis, resulting in a four-dimensional structure with the remaining 19 items.

1. Identity Safety,

2. Transport Safety,

3. Physical Safety,

4. Treatment/Nutritional Safety.

The sub-dimensions of the scale were structured based on the themes and sub-dimensions obtained in the qualitative phase, in accordance with the explanatory sequential mixed methods approach, and within the conceptual framework. The scale aims to holistically assess the fundamental risk areas that newborns may face in disaster situations, including identity security, transport security, physical security, and treatment/nutrition security. Due to the scale’s dichotomous structure and its development for rapid assessment in disaster settings, item-total correlations and content-based evidence were primarily considered in the structural validity assessment. This structure allows for the early and systematic identification of preventable risks related to identity errors, transfer process risks, environmental and care-related harms, and treatment/nutrition practices. In this respect, it is believed that the scale can fill a significant gap in the literature by enabling a structured and rapid assessment of newborn safety in disaster conditions. Furthermore, it is anticipated that future factor analyses using a tetrachoric correlation matrix with larger samples will further strengthen the structural validity of the scale.

The results of the present study indicate that the instrument developed for evaluating neonatal safety in disaster settings is a valid and reliable tool. The findings provide robust empirical evidence supporting the validity, reliability, and clinical applicability of the NSASD.

Conclusion

In this study, it was concluded that the ‘Neonatal Safety Assessment Scale in Disasters’, developed to evaluate the safety of newborns in disasters, is a valid and reliable instrument that can be effectively utilized by pediatric nurses providing neonatal care. Regarding the safety score obtained by summing the scale items, higher scores indicate greater safety, whereas lower scores indicate greater risk. The scale comprises four sub-dimensions covering neonatal risks: Identity Safety, Transport Safety, Physical Safety, and Treatment/Nutritional Safety. The NSASD is a valid, reliable, and practical tool for assessing neonatal safety in disaster contexts.

Limitation and strengths

A limitation of this study is that it was conducted with nurses following the 2023 earthquake in Turkey; therefore, the findings may not generalize to other types of disasters. Although the dichotomous response format of the scale facilitates ease of use and rapid decision-making, it imposes certain constraints from an analytical perspective. The dichotomous data structure may lead to consequences such as failing to meet parametric analysis assumptions, limited applicability of factor analyses, and reduced statistical power due to low item variance. Since Likert-type scales in the literature allow for broader comparisons, the limited number of studies available for comparison with similar measurement instruments represents one of the challenges of this study. The fact that data was collected using both online and in-person methods, and the differences in application times, should be considered a potential limitation that may have affected the response processes. However, the same scale and standard guidelines were used in both methods. Another limitation of the study is that it was conducted solely with nurses; the exclusion of other healthcare professionals (physicians, paramedics, etc.) may limit the generalizability of the findings. Furthermore, the lack of any assessment regarding the evacuation status of newborns can also be considered a shortcoming. Another limitation is that the study relies on nurses’ self-reported data; this may lead to response bias.

The data obtained in the qualitative phase of the research, conducted using a sequential mixed-methods design, align with the risks defined in the literature and reflect current field experiences. The scale items were generated based on a literature review, expert opinions, and field observations following the major earthquake that occurred in Turkey on February 6, 2023. In this respect, the study presents an original instrument shaped not only by theoretical frameworks but also by evidence based on real disaster experiences. Its dichotomous structure makes it easy for practitioners to use. By covering four fundamental safety domains (identity, physical, transport, and treatment/nutritional safety), the scale enables systematic, multidimensional assessment of newborns in disaster situations. The scale makes a significant contribution regarding prioritization, standardization of care, and early risk identification in post-disaster healthcare services. These findings strongly support the validity and reliability of the NSASD. Applying the study in the context of a real-world disaster, such as the 2023 Türkiye earthquake, plays a significant role in revealing the effectiveness obtained. The use of a mixed methods approach strengthens the validity of the scale. The practical value provided by the dichotomous response format in crisis situations supports ease of application and rapid decision-making.

Author contributions

This research was conducted as a thesis within the scope of the Pediatric Nursing master’s program at Hatay Mustafa Kemal University. All stages of the research were reviewed and conducted jointly by the advisor and the author. Assoc. Prof. Emel Demir designed the study, constructed the dataset, and drafted parts of the initial manuscript. Res. Assist. Rana Rehap Sert collected the data and performed the statistical analyses. All authors reviewed and approved the final version of the manuscript and take full responsibility for its content.

Funding

The Mustela Foundation financially supported this study.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to ethical considerations and participant confidentiality. Still, they are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.