Abstract

Nurses play a key role in health emergency preparedness; however, there is insufficient literature related to nurse disaster readiness in resource-limited and high-risk environments. In GB, which is considered one of the most disaster-prone locations in South Asia, there is no empirically validated scale measuring DNCC. In this paper, a cross-case analysis has been conducted to determine if combinations of Education (EDU), Training (TRN), Organizational Support (ORG) and Structural Barriers (BAR) are necessary and sufficient to generate high or low levels of DNCC in nurses working at public hospitals. Fuzzy-set Qualitative Comparative Analysis (fsQCA) was used for data analysis using direct calibration based on percentile. It has been observed that none of the above-mentioned factors alone are sufficient to generate high DNCC. However, combination of different enabling factors can be responsible for high DNCC, while no organizational support has consistently shown up in configurations leading to low DNCC.

Keywords

Core Competencies in Disaster Nursing, Fuzzy-set QCA, Configurational Analysis, Causal Asymmetry, Nursing Education, Organizational Support, Structural Barriers, Gilgit-Baltistan, Pakistan, Sendai Framework.

Introduction

Nursing is the largest professional group in any health system and the largest professional group to care during mass casualty events. Their performance in such situations does not depend upon seniority or overall clinical experience but rather on a specific, learnable set of disaster nursing core competencies (DNCC): triage proficiency, incident command literacy, crisis communication, decontamination procedures, psychological first aid and ethical resource allocation (ICN & WHO, 2009). Systematic reviews consistently report low-to-moderate DNCC across the world with the most severe shortages concentrated in the areas of highest exposure to disasters (Labrague et al., 2018; Heng et al., 2022). Gilgit-Baltistan (GB) is the symbol of this disparity. It is located on the intersection of the Karakoram, Himalayan, and Hindu Kush ranges, and includes more glaciers than any non-polar landmass and faces compounding, exacerbating hazards: glacial lake outburst floods (GLOFs), seismic events, slope failures, and seasonal inundation (Farhan et al., 2020; Rasul et al., 2019). Super-floods of 2022 led to the destruction of health facilities in all five districts. But the nursing staff that needs to react to such incidents had never been empirically evaluated in the context of DNCC a gap that informed the current study. It also lacks a methodological gap. The use of multiple regression is virtually exclusive in the DNCC studies in Pakistan and similar low- and middle-income countries (LMICs). These assumptions are misleading and reflective of preparedness phenomena. A nurse working in a high-barrier environment in which she is not supported by her organization and whose disaster education has prepared her to work in an environment with high barriers to entry but low barriers to exit does not occupy some intermediate position on a competency continuum she occupies a qualitatively distinct institutional structure with its own causal properties. This cannot be identified by regression, fuzzy-set Qualitative Comparative Analysis (fsQCA) is made to do so. Set-theoretic causation is operationalized by the set-theoretic causal recipe (fsQCA) (Ragin, 2008; Schneider & Wagemann, 2012), the set of conditions that jointly and only together suffice to cause an outcome. It supports equifinality (more than one way to get to the same place), causal asymmetry (structurally distinct enabling and inhibiting configurations), and INUS causation (conditions insufficient alone but necessary within a sufficient combination). Using fsQCA on the data of 312 nurses in five DHQ hospitals in GB, the study posits not which factor is significant. most but which configurations of Education (EDU), Training (TRN), Organizational Support (ORG), and the absence of Structural Barriers (~BAR) are sufficient for high and low DNCC.

Literature Review

The ICN framework (ICN & WHO, 2009) delineates ten DNCC domains spanning triage, decontamination, incident command, psychological first aid, and ethical resource allocation across all four disaster management phases. Irrespective of this established taxonomy, nurses around the world self-report moderate-to-low competency (Labrague et al., 2018), with specific deficiencies in managing the surge capacity, understanding the disaster plan, and being more precise with the triage (Ghanbari et al., 2020). Heng et al. (2022) explained these gaps mainly by structural factors: insufficient curricula and a lack of institutional mechanisms. Pakistan Disaster-specific content in most provincial nursing programmes is not present, and GB institutions do not exist at all in either of the national curriculum audit reports conducted to date (Aqil et al., 2023; Majeed et al., 2021). Disaster-specific education instead of general level of qualification is the one that is enhancing DNCC. In Hong Kong, Fung et al. (2021) demonstrated that nurses who had completed dedicated disaster modules were significantly more effective than nurses with the same general credentials. Shabbir et al. (2022) earlier replicated the same study in a Pakistani sample. Practical training operationalizes cognitive knowledge under controlled stress: In Alzahrani and Kyratsis (2020) study, substantially higher competency retention was found in simulation-trained nurses, and Lateef et al. (2021) confirmed that participation in drills within the 12 months prior to the study was an independent predictor of multi-domain DNCC proficiency. Organizational support including involvement of nurses in disaster planning, authority to activate emergency protocols, institutional training provision, and special disaster nursing roles have become unique enabling dimensions in recent literature. Zhang et al. (2021) observed that the higher DNCC in Chinese hospitals with identified disaster nursing coordinators compared to those without such coordinators. Organizational factors were ranked as the strongest drivers of DNCC variation in an international review across a multi-site review (Al Thobaity et al., 2022). This paper assumes that ORG is an independent analytical condition, independent of barriers, to gain its independent enabling role. Barriers to DNCC work at various levels: structural (lack of curricula, limited access to continuing education in conservative settings), organizational (no expert personnel, no institutional plans to mitigate disasters), and individual (participation barriers in conservative settings are gender-related). The strongest suppressors of DNCC across settings were found to be barriers by Al Thobaity et al. (2022). GB reinforces these pressures with geographic isolation, chronically under-invested in public hospitals, and institutional cultures where disaster preparedness has long been viewed as peripheral to core nursing practice.

Regression-based DNCC research imposes additive, symmetric assumptions that configurational complexity violates. If education's effect on DNCC is contingent on barrier levels and organizational context  if it is INUS-causal (Mackie, 1965; Ragin, 2008)  regression will consistently underestimate or misrepresent it. fsQCA has been productively applied in adjacent health workforce domains including patient safety (Dückers et al., 2019) and occupational burnout (Ramos et al., 2021), but has not, to our knowledge, been applied to DNCC. The present study fills this gap.

Conceptual Framework

The study draws on two established frameworks. The ICN Framework of Disaster Nursing Competencies (ICN & WHO, 2009) provides the theoretical structure for the outcome variable across all four disaster phases. Sendai Framework Priority 4 building health workforce preparedness at national and sub-national levels furnishes the policy rationale, situating this study as a contribution to Pakistan's Sendai implementation evidence base.

The four conditions are conceptualized at distinct system levels: EDU at the curriculum level, TRN at the facility level, ORG at the institutional governance level, and BAR at the structural and contextual level. These variables do not operate separately; rather, they are presumed to operate in such a way that their combined presence or absence will result in different preparedness states, which cannot be represented by additivity models. In fact, barriers are proposed as being a blocking variable – a variable whose presence prevents the translation of input variables into competencies, regardless of the level of the inputs. Here, competencies emerge from the interaction of conditions, not as an aggregation of inputs.

Methodology

Data were collected between October 2025 and February 2026 at five DHQ hospitals: DHQ Gilgit (Gilgit District, n = 78), DHQ Skardu (Baltistan, n = 72), DHQ Aliabad (Hunza-Nagar, n = 56), DHQ Gahkuch (Ghizer, n = 54), and DHQ Chilas (Diamer, n = 52). These facilities serve as designated mass casualty response centres for their respective districts and together cover the largest share of GB's population. All nursing staff available across all three daily shifts during the collection period were invited to participate. Of 338 eligible nurses, 312 completed the questionnaire (response rate: 92.3%); non-response was due to leave, shift absence, or personal refusal, with no evidence of systematic demographic bias.

fsQCA was chosen because the research question concerns which combinations of conditions are sufficient for preparedness a set-theoretic rather than a net-effects question, and one for which regression is an ill-suited tool. The questionnaire was adapted from the Disaster Nursing Core Competency Scale (Al Thobaity et al., 2016), which has been validated across South and Southeast Asian contexts. The 48-item instrument measured DNCC (29 items), Barriers (8 items), Education (4 items), Training (3 items), and Organizational Support (4 items) on a five-point Likert scale (1 = Strongly Disagree; 5 = Strongly Agree). It was piloted with 20 nurses at a non-study hospital in Rawalpindi before field deployment. Internal consistency estimates are reported in Table 1. Calibration converts continuous subscale means into fuzzy membership scores ranging from 0 (full non-membership) to 1 (full membership), with 0.5 marking the crossover point of maximum ambiguity. Three qualitative anchors are required under the direct method (Ragin, 2008): full membership (fm), crossover (cp), and full non-membership (fnm).

Review of the data indicated that there was extreme positive skewness in all three subscales of the enabling conditions (EDU M = 1.91; TRN M = 1.79; ORG M = 1.64). Adopting the standard values for Likert-based anchors (fm = 4.0, cp = 3.0, fnm = 2.0) would have resulted in more than 97% of the observations being rated as completely non-members of the conditions, which would render them meaningless to analysis. Thus, percentile-based anchors were used, whereby fm is the 95th percentile, cp is the median, and fnm is the 5th percentile. This approach is consistent with the advice offered by Ragin (2008, p. 91). Barrier thresholds were applied in inverted order so that high barrier scores correspond to low ~BAR (absence of barriers) membership. The logistic transformation function was applied throughout. All thresholds and resulting descriptive statistics are reported in Table 1.

Table 1

Instrument Reliability, Descriptive Statistics, and Calibration Thresholds (N = 312)

Variable	?	Items	M	SD	Min	Max	fm (P95)	cp (P50)	fnm (P05)
DNCC (Outcome)	.967	29	2.99	0.61	1.45	4.59	4.03	3.00	2.05
Education (EDU)	.787	4	1.91	0.59	1.00	4.00	3.00	1.75	1.00
Training (TRN)	.621	3	1.79	0.45	1.00	3.00	2.67	1.67	1.00
Org. Support (ORG)	.712	4	1.64	0.46	1.00	3.25	2.50	1.50	1.00
Barriers (BAR)	.885	8	4.02	0.55	2.25	5.00	3.12†	4.00†	4.88†

Note: ? = Cronbach's alpha computed from item-level responses. fm = full membership threshold (P95); cp = crossover point (P50/median); fnm = full non-membership threshold (P05). † BAR thresholds are expressed in terms of the original BAR scale and applied in inverted order for the ~BAR condition: a low BAR score corresponds to high ~BAR membership. Full-scale ? = .920. The TRN ? of .621 reflects a 3-item subscale; this condition should be interpreted with appropriate caution.

Necessity analysis examines whether any single condition is consistently present whenever the outcome occurs set-theoretically, Y ? X. The threshold for necessity is consistency ? 0.90 (Schneider & Wagemann, 2012). The Relevance of Necessity (RoN) was also computed: values below 0.60 indicate that a condition is present so broadly across all cases as to be analytically uninformative. Necessity was tested for both the high- and low-DNCC outcomes.

Truth Table and Boolean Minimization

Assignment to the conditions was done on the basis of whether fuzzy score was more than 0.5 (cases that scored above 0.5 were considered as members, while below 0.5 were classified as non-members). Any case close ±0.05 to the crossover point in terms of any one condition was excluded because of ambiguity, leading to a reduction of the sample size used to 149 out of 16 possible configurations in the truth table, of which only 14 were empirically applicable, while the rest were taken as logical remainders. The Proportional Reduction of Inconsistency (PRI) measure was also computed for each configuration to rule out subset relations. Consistency of above 0.75 combined with PRI of over 0.60 were necessary criteria for sufficiency.

Solutions generated using Boolean minimization (Quine-McCluskey algorithm) included complex, parsimonious, and intermediate. As per Fiss (2011), the latter is provided as the main outcome since it is founded on theoretical assumption concerning the logic of remainders.

Causal Asymmetry and Robustness

Another truth table analysis under similar conditions but now for low DNCC (~DNCC) configurations was carried out. Causal asymmetry is established when there is a difference in structure between these ~DNCC configurations and their negations. In order to assess robustness, calibration thresholds were altered by ±0.25 units while performing a full analysis. The consistency threshold was increased to 0.80 for assessing structural robustness.

Ethics

Ethical approval was sought from SZABIST University Islamabad Research Ethics Committee (Ref: SZABIST-REC/2025/044) and Gilgit-Baltistan Health Department. Written consent was obtained from all respondents. The study was voluntary, anonymous, and respondents could withdraw from the study at any point without any implications.

Results

Sample Characteristics

Sample size included 164 women nurses (52.6%) and 148 men nurses (47.4%) who were divided into groups of LHVs (n = 116, 37.2%), SMTs (n = 97, 31.1%), JMts (n = 89, 28.5%), and RNs (n = 10, 3.2%). Over half of the nurses had Intermediate or Diploma degrees (n = 204, 65.4%), whereas 9.3% had a degree above or at bachelor's level. Over two-thirds of the nurses were permanently employed (n = 238, 76.3%). Work experience varied between 1-3 years (18.9%) and more than 10 years (27.2%). All five hospitals failed to conduct any disaster drills within the previous three years; furthermore, they did not

Descriptive Statistics, Correlations, and Regression

Table 2 presents bivariate correlations. BAR showed the strongest association with DNCC (r = ?.554, p < .001), followed by EDU (r = .421, p < .001), ORG (r = .165, p = .004), and TRN (r = .113, p = .046). Inter-predictor correlations were uniformly low (|r| ? .126), excluding multicollinearity as a concern. The regression model was statistically significant (R² = .453; Adjusted R² = .446; F(4, 307) = 63.54, p < .001). BAR was the dominant predictor (? = ?.505, p < .001), with EDU (? = .341, p < .001) and ORG (? = .126, p = .003) as significant positive predictors. TRN did not reach conventional significance (? = .081, p = .057), most likely because its scores were severely compressed across a sample in which no nurse had received structured disaster training in the preceding three years (M = 1.79, SD = 0.45, maximum observed = 3.00).

Table 2

Pearson Correlations and Multiple Regression Results (N = 312)

Variable	1. DNCC	2. BAR	3. EDU	4. TRN	5. ORG	?
1. DNCC (Outcome)
2. BAR (Barriers)	?.554**					?.505**
3. EDU (Education)	.421**	?.126*				.341**
4. TRN (Training)	.113*	?.051	.038			.081†
5. ORG (Org. Support)	.165**	?.017	.102	?.052		.126**
Model fit	R² = .453	Adj. R² = .446	F(4,307) = 63.54	p < .001

Note: ** p < .01; * p < .05; † p = .057 (marginal; below conventional significance threshold). Correlations derived from item-level subscale means. ? = standardized coefficient from the four-predictor model. DNCC = Disaster Nursing Core Competencies; BAR = Barriers; EDU = Education; TRN = Training; ORG = Organizational Support.

Gender and Experience Differences

Male nurses scored significantly higher on DNCC (M = 3.25, SD = 0.58) than female colleagues (M = 2.75, SD = 0.53; t(310) = 7.956, p < .001) and reported greater disaster education exposure (t = 6.77, p < .001). Female nurses faced substantially higher barriers (M = 4.17 vs. M = 3.85; t = ?5.32, p < .001), reflecting documented mobility and participation constraints in GB's social context. Training scores did not differ by gender (t = 0.867, p = .386), which indicates that institutionally delivered training, when provided, reaches both genders equitably an important consideration for program design. Work experience showed no relationship with DNCC: nurses with 1–3 years (M = 3.03, SD = 0.60) and those with over 10 years (M = 2.95, SD = 0.60) were statistically indistinguishable (t(142) = 0.790, p = .431). GCE, acquired in absence of disaster content, does not have any competencies over structured disaster content.

Necessity Analysis

The table for necessity findings is included in Table 3. Not one of the conditions exceeded the ? 0.90 necessity criteria for a high level of DNCC. The highest values of consistency are observed in case of EDU (0.788), and ~BAR (0.778). It should be noted that it corresponds to equifinality nature of causation. Since there are various ways to achieve a high level of DNCC, each enabling condition will not be available in each high-competent person. RoN findings (0.474 in case of EDU, 0.492 in case of ~BAR) show that these results are not a coincidence.

When analyzing low DNCC, the necessity criterion was almost achieved: BAR (0.800) had a maximum value among other values of consistency. In other words, barriers are found in almost each case when competency level is low.

Table 3

Necessity Analysis for High DNCC and Low DNCC (N = 312)

Condition	Cons. (High DNCC)	Cov.	RoN	Cons. (Low DNCC)	Cov.	Assessment
EDU	0.788	0.728	0.474	0.600	0.588
TRN	0.722	0.651	0.711	0.679	0.648
ORG	0.720	0.663	0.687	0.639	0.624
~BAR	0.778	0.786	0.492	0.509	0.544
BAR	0.549	0.513	2.034	0.800	0.793	Near-nec.†
~EDU	0.554	0.567	2.109	0.722	0.783
~TRN	0.609	0.642	1.407	0.634	0.708
~ORG	0.592	0.607	1.456	0.655	0.712

Note. Necessity threshold: consistency ? 0.90 (Schneider & Wagemann, 2012). Condition does not obtain necessity for high DNCC. RoN = Relevance of Necessity; values greater than 2.0 indicate trivial necessity. † The BAR is approaching necessity for low DNCC (consistency = 0.800; less than the strict threshold but nonetheless highly relevant given the structural role it plays in inhibition). ~ = logical negation. Cons. = consistency; Cov. = coverage.

Truth Table

Table 4 shows the complete truth table. Of the total 16 configurations that can theoretically exist, 14 configurations have been filled in empirically, while the remaining 2 configurations do not contain any observations and have therefore been treated as logical remainders. Out of 14 configurations, 5 configurations met both the criteria (consistent score > 0.75 and PRI > 0.60), namely, 0111, 1001, 1011, 1101, and 1111.

Table 4

Truth Table: Configurations for High DNCC (Unambiguous Cases n = 149 of 312)

Config.	EDU	TRN	ORG	~BAR	n	Consistency	PRI	Outcome	Type
0000	?	?	?	?	7	0.185	0.000	0
0001	?	?	?	?	3	0.610	0.066	0
0010	?	?	?	?	5	0.266	0.000	0
0011	?	?	?	?	9	0.712	0.673	0
0100	?	?	?	?	7	0.225	0.000	0
0101	?	?	?	?	9	0.713	0.867	0
0110	?	?	?	?	16	0.381	0.535	0
0111	?	?	?	?	11	0.801	0.884	?	Sufficient
1000	?	?	?	?	11	0.619	0.662	0
1001	?	?	?	?	3	0.916	0.892	?	Sufficient
1010	?	?	?	?	9	0.468	0.672	0
1011	?	?	?	?	7	0.955	0.965	?	Sufficient
1100	?	?	?	?	9	0.733	0.727	0
1101	?	?	?	?	8	0.986	0.980	?	Sufficient
1110	?	?	?	?	20	0.607	0.857	0
1111	?	?	?	?	15	0.843	0.953	?	Sufficient

Note:  condition present (fuzzy score > 0.5); ? = condition absent. If consistency is greater than 0.75 and PRI is greater than 0.60, outcome coding was 1. N= number of cases in the subset truth table. There were two instances with no empirical cases. PRI = Proportional Reduction in Inconsistency.

Sufficient Configurations for High DNCC

The Boolean simplification resulted in three paths (see Table 5; Solution Consistency = 0.885; Solution Coverage = 0.645). The optimal path simplifies to:

High DNCC  ?  (EDU * ~BAR)  +  (TRN * ORG * ~BAR)

Barrier absence (~BAR) appears in both parsimonious terms, confirming its structural centrality. The intermediate solution restores peripheral conditions, as shown in Table 5.

Table 5

fsQCA Intermediate Solution: Sufficient Configurations for High DNCC

Path	EDU	TRN	ORG	~BAR	Boolean Expression	Raw Cov.	Unique Cov.	Consistency
P1	?	?	?	?	EDU * TRN * ~BAR	0.487	0.111	0.910
P2	?	?	?	?	EDU * ORG * ~BAR	0.483	0.105	0.914
P3	?	?	?	?	TRN * ORG * ~BAR	0.444	0.072	0.881
SOLUTION						Cov. = 0.645		Cons. = 0.885

Note: condition present (core if retained in parsimonious solution; peripheral if intermediate only); ? = condition absent. * = Boolean AND. EDU and ~BAR are core in P1 and P2; TRN, ORG, and ~BAR are core in P3. Raw Cov. = raw coverage; Unique Cov. = unique coverage; Cons. = consistency. Solution type: intermediate.

P1 (EDU*TRN*~BAR)

Education and training co-occur in a low-barrier environment, which results in high DNCC (consistency = 0.910; raw coverage = 0.487). Peripheral here is its absence which is accepted during the minimization process, but P2 shows that it can substitute training when it is present. P1 best explains those nurses who got well-organized disaster material via pre-service or in-service program and work in an environment where barriers to access are minimal.

P2 (EDU*ORG*~BAR)

Organizational support replaces the missing training (consistency = 0.914; raw coverage = 0.483). In place of regular exercises, the educational material becomes competency in the presence of meaningful hospital governance in nurses involved in the planning, empowered to act, and operating within delimited disaster roles. The almost identical raw coverage of P1, combined with a small amount of unique coverage (0.105), suggests that there is a large amount of empirical overlap between these two pathways.

P3 (TRN*ORG*~BAR)

It is a combination of training and organizational support that generates high DNCC without formal disaster education (consistency = 0.881; raw coverage = 0.444). This line of reasoning demonstrates that institutional scaffolding and experience can address gaps in the curriculum. The fact that its unique coverage is relatively low (0.072) implies that the majority of P3 cases are also covered by P1 or P2.

Sufficient Configurations for Low DNCC

Table 6 presents the intermediate solution for low DNCC (solution consistency = 0.918; coverage = 0.513). Both pathways share ~ORG*BAR as their core  the irreducible inhibiting kernel is absent organizational support within a high-barrier environment. N1 adds absent education (~EDU*~ORG*BAR; consistency = 0.954); N2 adds absent training (~TRN*~ORG*BAR; consistency = 0.923). Importantly, N2 does not presuppose the absence of education: a nurse who has been taught about the disaster and has not been trained, supported in the organization and with high structural barriers is still on the way to low competency. The finding directly refutes any policy supposition that curriculum reform per se is a sufficient preparedness intervention.

Table 6

fsQCA Intermediate Solution: Sufficient Configurations for Low DNCC (~DNCC)

Path	~EDU	~TRN	~ORG	BAR	Boolean Expression	Raw Cov.	Unique Cov.	Consistency
N1	?	?	?	?	~EDU * ~ORG * BAR	0.447	0.098	0.954
N2	?	?	?	?	~TRN * ~ORG * BAR	0.424	0.077	0.923
SOLUTION						Cov. = 0.513		Cons. = 0.918

Note: condition present (core if also in parsimonious solution); ? = condition absent. ~ORG and BAR are core in both pathways. ~DNCC = low disaster nursing core competencies. Causal asymmetry is confirmed: enabling (Table 5) and inhibiting (Table 6) configurations are structurally distinct and are not logical mirrors of one another.

Robustness

Calibration thresholds (±0.25 units) with the alternative calibration thresholds yielded substantively identical core pathways with only peripheral conditions varying. By increasing the cutoff value of consistency to 0.80, the configuration 1111 was removed in the sufficient set, however, the parsimonious solution was retained as well as the structure of the inhibiting pathway. The intermediate is resistant to moderate changes in specification.

Discussion

Preparedness Is Configurational, Not Additive

The main findings of the research is the fact that high DNCC in GB is the result of certain combinations of the enabling factors, but not of gradual increases of any of the enabling factors. Three pathways yield high competency with each one needing the absence of barriers and at least two enabling conditions with none requiring all three simultaneously. It is the limitation of additive regression models that by design cannot reveal that the effect of education on competency is conditional on the level of barriers or organizational setting. It is not that regression identifies the wrong predictors it uses the wrong causal model.

This finding is consistent with Blanchet et al.'s (2017) argument that health workforce capacities emerge from system-level configurations rather than from discrete training investments. The present study provides the first empirical test of this claim in a disaster nursing context.

The Structural Centrality of Barriers

Inhibiting configurations contain barriers and enabling ones do not. Though the structural role of ~BAR was short of strict necessity (consistency = 0.778), the structural role of the same is shared by all three pathways to high DNCC and core in both inhibiting pathways and thus makes it the most policy-consequential finding to emerge in this analysis. The suppressive effect of barriers as a linear quantity is captured in the regression coefficient ( = -.505); something more specific, barriers are a blocking condition, is revealed in the structure of the fsQCA. These barriers in GB are not personal but institutional. Geographic distance with ongoing education, the lack of disaster nursing knowledge in hospital hierarchies, and especially in the case of female nurses constraints of participation due to the social nature of the region is all characteristics of the system. They elaborate why work experience is not associated with any competency benefit: a 10-year general clinical practice experience in a high-barrier, zero-training environment does not contribute any competency benefit to disaster-specific preparedness (t = 0.790, p = .431). Organizational Support: Compensatory and Asymmetric. Regression finds ORG to be a significant predictor ( =.126), but the configurational analysis suggests a more textured role. In P2, organizational support replaces training absent hospitals with substantial governance enable the educational material to transform into competency without routine drills. In P3, the absence of ORG is replaced by training. However in both inhibiting pathways, ~ORG is core: the lack of organizational support is a predictable contributor to low DNCC despite how other conditions might vary. This asymmetry where the lack of a certain aspect causes harm is more consistently true than the presence of a certain aspect causes benefit: there must be a minimum governing aspect of an organization before enabling conditions can work. The 5 study hospitals were all below the ORG crossover point (grand M = 1.64 on a 15 scale). None had any disaster nursing coordinator; none had a nurse-inclusive disaster planning committee; none had conducted a drill in the last three years. The practical implication is obvious: assigning one trained coordinator per hospital is the cheapest, highest leverage structural change available and would start the enabling dynamic that is demonstrated in P2 and P3 even before the curriculum reforms can be implemented.

Causal Asymmetry and Its Policy Corollary

The inhibiting configurations are structurally simpler than the enabling ones: both require only ~ORG*BAR as their core, with one additional absent enabling condition. This asymmetry has a practical corollary that regression cannot produce: disrupting low-DNCC configurations demands a smaller intervention bundle than building high-DNCC ones. Establishing minimal organizational support and reducing the most tractable barriers may be sufficient to break the inhibiting pathway even before formal education or training programs are in place. It can be seen through causal asymmetry only.

Training's Marginal Regression Coefficient Does Not Signal Irrelevance

The no significance of TRN in the regression model (? = .081, p = .057) appears more likely due to restriction of range than ineffectiveness. The scores have been constricted in the context of a sample in which none of the nurses have undergone training in managing disasters over the past three years (M = 1.79, SD = 0.45; highest score obtained = 3.00). The fsQCA structure is more informative: TRN is core in P1 and P3, and its absence from P2 reflects substitutability by organizational governance, not irrelevance. Studies sampling from settings where training exposure is more variable will be needed to clarify this point.

Alignment with the Sendai Framework

Sendai Framework Priority 4 requires investment in health workforce preparedness at sub-national levels. Pakistan's Sendai National Action Plan designates workforce training as a primary mechanism. The present findings challenge this emphasis in isolation: training investment without complementary barrier reduction and organizational governance will not produce preparedness gains in settings like GB. The configurational evidence points instead to simultaneous action addressing barriers, establishing governance, deploying education, and providing training concurrently rather than in sequence.

Policy Implications:

Prioritize barrier reduction

Before deploying training programmes or curriculum reforms, the GB Health Department and GBDMA should conduct a systematic barrier audit across all DHQ hospitals. Removing the most tractable barriers restricted continuing education access, gender-exclusive participation norms, and the institutional invisibility of disaster nursing as a professional role is a prerequisite for any other intervention to take effect.

Designate disaster nursing coordinators

A single trained coordinator per hospital would cross the ORG threshold identified in this analysis, activating the enabling roles demonstrated in P2 and P3 at minimal cost. This is the most accessible structural change available to hospital management in the near term.

Reform curriculum alongside not before institutional investment

The Pakistan Nursing Council's mandate to integrate disaster nursing into diploma and degree programmes is necessary but not sufficient. Curriculum reform must be pursued concurrently with facility-level training infrastructure and organizational governance development. The N2 pathway makes clear that education, in isolation, cannot produce preparedness in GB's current institutional environment.

Build training programmes that are gender-equitable by design

The absence of a gender difference in training access (p = .386) when training is institutionally delivered confirms that facility-based provision is an effective equalizer. Scheduling across all shifts with explicit inclusion provisions will maintain this equity as programmes expand.

Limitations

The cross-sectional design does not support temporal causal inference; the configurations identified reflect correlational structures that fsQCA interprets causally but that require longitudinal validation. All data are self-reported; anonymous administration and strong full-scale reliability (? = .920) reduce but do not eliminate social desirability concerns. The TRN subscale comprises only three items (? = .621), and caution is warranted when interpreting this condition in isolation; future research should employ longer subscales. Calibration thresholds involve researcher judgment, though robustness checks confirmed core pathway stability across a ±0.25-unit shift in all thresholds. The sample is confined to public hospital nurses in GB; community health workers, private-sector nurses, and nursing students were not included, and the configurational findings may not transfer to higher-resource settings in which some enabling conditions are already in place. Several potentially relevant conditions prior disaster experience, team cohesion, personal motivation were not modeled and may shape configurations that the present analysis does not capture. Finally, the ORG subscale (4 items, ? = .712) warrants replication with a purpose-built validated instrument.

Conclusion

This study provides the first empirical assessment of DNCC in Gilgit-Baltistan and the first application of fsQCA to disaster nursing preparedness research in Pakistan. Its central argument is that preparedness is not the aggregate of its determinants but the product of their configuration. No single enabling condition is either necessary or sufficient. High DNCC requires at least two concurrent enabling conditions within a structurally low-barrier environment, and multiple distinct pathways can achieve this. The inhibiting configurations are structurally simpler absent organizational support combined with high barriers reliably produces low competency regardless of educational level and disrupting them requires a smaller intervention bundle than producing high competency.

For the nurses of GB, this evidence is both diagnostic and actionable. It is simply not prepared enough to deal with the problems it encounters, not due to lack of initiative or ability on the part of nurses, but rather because they exist in organizational settings which inhibit their development of competence. This solution should itself take on a configurational form by addressing the problem through barrier reduction, organizational governance, training, and education.

References