Premature mortality due to cervical cancer: study of interrupted time series

Nascimento, Maria Isabel do; Massahud, Felipe Corrêa; Barbosa, Nathália Giácomo; Lopes, Cássio Destefani; Rodrigues, Vanessa da Costa

doi:10.11606/s1518-8787.2020054002528

Acessibilidade / Reportar erro

Brasil

Español English

sumário « anterior atual seguinte »

Sumário

Original Article • Rev. Saúde Pública 54 • 2020 • https://doi.org/10.11606/s1518-8787.2020054002528 copy

Premature mortality due to cervical cancer: study of interrupted time series

Authorship SCIMAGO INSTITUTIONS RANKINGS

ABSTRACT

OBJECTIVE:

To verify the effect of the Pact for Health on premature mortality (30–69 years) attributed to cervical cancer in Brazil and its macroregions, using interrupted time series analysis.

METHODS:

Segmented regression was used to assess “change in level” and “change in trend” in premature mortality rates attributed to cervical cancer considering the post-Pact period (2010-2018), controlling by the pre-Pact period (1998–2006). Understanding the triennium 2007-2009 as essential for the adoption and implementation of the policy, it was excluded from the main modeling, but assessed in the sensitivity analysis.

RESULTS:

From 1998 to 2018, there were more than 119,000 deaths due to cervical cancer in women aged 30 to 69 years in Brazil. The Northern region experienced the highest rates (> 20 per 100,000). Comparing with baseline (1998–2006), segmented regression showed a progressive increase in changing trend from cervical cancer deaths in Brazil as a whole (coefficient = 0.513; 95%CI 0.430 to 0.596) and in the Southeast region (coefficient = 0.515; 95%CI 0.358 to 0.674), South region (coefficient = 0.925; 95%CI 0.642 to 1.208), and Midwest region (coefficient = 0.590; 95%CI 0.103 to 1.077). The Northeast region presented the most promising effects with immediate reduction in change level (-0.635; 95%CI −1.177 to −0.092) and progressive reduction in the changing trend of premature deaths (coefficient= −0.151; 95%CI −0.231 to −0.007).

CONCLUSIONS:

Premature mortality rates due to cervical cancer are high in Brazil and its macroregions. This interrupted time series was not able to reveal the effectiveness of initiatives related to the Pact for Health on premature deaths from cervical cancer nationally and in all macroregions equally. The best results are restricted to the Northeast region.

DESCRIPTORS:
Public Policy; Noncommunicable Diseases; Uterine Cervical Neoplasms; Mortality, Premature; Interrupted Time Series Analysis

Final behavior of mortality after intervention			Initial exploratory model with no serial autocorrelation terms				Final model with n terms of serial autocorrelation
Regressors		Interpretation	Coefficient	95%CI	DW Statistic ^a a Durbin Watson statistic (p value < 0.05 suggests serial autocorrelation).	AIC	Coefficient	95%CI	AIC ^b b Akaike Information Criterion (smaller values suggest better quality of fit).	Terms AR ^c c Terms of autocorrelation defined by the function of autocorrelation and partial autocorrelation.
Brazil
	Change in level	Abrupt elevation	0.48	-0.467 to 1.428	0.086	36.772	0.497	0.039 to 0.955	15.918	p = 3
	Change in trend	Progressive elevation	0.495	0.351 to 0.638			0.513	0.430 to 0.596
	Intercept		18.872	18.301 to 19.443			18.821	18.518 to 19.124
	Time		-0.460	-0.561 to −0.358			-0.467	-0.523 to −0.411
North
	Change in level	Abrupt elevation	2.522	-0.086 to 5.131	NS	65.065	2.629	0.422 to 4.837	61.234	p = 1
	Change in trend	Not detected	0.118	-0.276 to 0.513			0.128	-0.207 to 0.464
	Intercept		22.703	21.132 to 24.274			22.725	21.399 to 24.052
	Time		-0.122	0.402 to 0.156			−0.133	0.371 to 0.104
Northeast
	Change in level	Abrupt reduction	-1.062	-2.149 to 0.025	0.018	40.561	-0.635	-1.177 to −0.092	26.596	p = 2
	Change in trend	Progressive reduction	-0.142	-0.307 to 0.022			-0.151	-0.231 to −0.071
	Intercept		14.494	13.839 to 15.149			14.516	14.201 to 14.833
	Time		0.079	0.036 to 0.196			0.058	0.0007 to 0.116
Southeast
	Change in level	Not detected	0.881	-0.122 to 1.886	0.098	38.336	0.889	-0.156 to 1.934	27.669	p = 1
	Change in trend	Progressive elevation	0.517	0.365 to 0.669			0.516	0.358 to 0.674
	Intercept		15.754	15.149 to 16.359			15.753	15.123 to 16.383
	Time		-0.609	-0.716 to −0.502			-0.609	-0.721 to −0.497
South
	Change in level	Not detected	1.051	-0.668 to 2.770	0.098	53.394	1.059	-0.821 to 2.940	46.765	p = 1
	Change in trend	Progressive elevation	0.918	0.657 to 1.178			0.925	0.642 to 1.208
	Intercept		20.091	19.056 to 21.127			20.131	18.995 to 21.265
	Time		-0.869	-1.053 to −0.685			-0.873	-1.073 to −0.673
Midwest
	Change in level	Not detected	0.994	-2.624 to 4.613	NS	74.232	0.933	-2.271 to 4.137	73.43	p = 1
	Change in trend	Progressive elevation	0.565	0.017 to 1.112			0.59	0.103 to 1.077
	Intercept		21.283	19.104 to 23.463			21.333	19.407 to 23.260
	Time		-0.734	-1.121 to −0.346			-0.743	-1.087 to −0.398

Faculdade de Saúde Pública da Universidade de São Paulo Avenida Dr. Arnaldo, 715, 01246-904 São Paulo SP Brazil, Tel./Fax: +55 11 3061-7985 - São Paulo - SP - Brazil
E-mail: revsp@usp.br

Acompanhe os números deste periódico no seu leitor de RSS

[1] Correspondence: Maria Isabel do Nascimento, Faculdade de Medicina/UFF Secretaria da Coordenação de Medicina Rua Marquês do Paraná, 303, 2° andar 24033-900 Niterói, RJ, Brasil E-mail: ysamaria@uol.com.br